Skip to main content
NCBI home page
UKPMC Funders Author Manuscripts logoLink to UKPMC Funders Author Manuscripts
. Author manuscript; available in PMC: 2026 Aug 11.
Published before final editing as: J Infect Dis. 2026 Mar 3:jiag093. doi: 10.1093/infdis/jiag093

Nasopharyngeal microbiome composition and its clinical correlates in children hospitalized with severe pneumonia in East Africa

Timothy O Makori 1,2,3,✉, Elijah T Gicheru 1,2, Maureen W Mburu 1,2, Mercy S Sada 1,2, Omar Nyawa 1,2, Martin Mutunga 1, Clement Lewa 1, Robinson Cheruiyot 1, Sarah Kiguli 4, Peter Olupot-Olupot 5, Rita Muhindo 5, Christabel Mogaka 6, Thomas N Williams 7,8, Charles N Agoti 1,2,6,7, Kathryn Maitland 7,8, Charles J Sande 1,2,6,7,✉
PMCID: PMC7619340  EMSID: EMS216166  PMID: 41773862

Abstract

Background

Pneumonia remains the leading cause of infectious mortality in children under five, with the highest burden in sub-Saharan Africa. Dysbiosis in nasopharyngeal (NP) microbiota may influence pneumonia susceptibility and progression, but little is known about its composition or clinical relevance in low- and middle-income countries (LMICs). We characterized the NP microbiota of children hospitalized with severe pneumonia in East Africa and investigated associations with clinical outcomes.

Methods

We performed 16S rRNA partial gene sequencing of NP swabs collected at hospital admission from 876 children enrolled in the COAST trial across five sites in Kenya and Uganda. Clinical, demographic, and virological data were prospectively collected. Microbial profiles were analysed using hierarchical clustering, non-metric multidimensional scaling (NMDS), and multivariable regression to assess associations with respiratory viral infections, sepsis, cyanosis, bacteraemia, coma, HIV status, malnutrition, sickle cell disease (SCD), malaria, and mortality.

Results

The nasopharyngeal microbiome was structured in six distinct clusters, each dominated by different genera, including Staphylococcus, Streptococcus, Haemophilus, Dolosigranulum, Corynebacterium, and Moraxella. Multivariable models adjusting for study site and age showed a positive association between Corynebacterium and early mortality. Temporal analysis showed elevated Corynebacterium abundance in children who died within 48 hours of admission, then declined over longer 56 survival intervals, approaching levels observed in survivors. However, time-continuous models did not support this persistent association, suggesting a subgroup effect.

Conclusion

We provide one of the largest high-resolution surveys of the paediatric upper airway microbiome in Africa, identifying microbial patterns associated with viral infection, HIV status, early death, and bacteraemia.

Keywords: Nasopharyngeal microbiome, Paediatric pneumonia, 16s rRNA sequencing

Introduction

Pneumonia is the leading infectious cause of death among children under five years of age, accounting for approximately 900,000 deaths globally, the majority of which occur in LMICs [1]. The mortality rate in these settings is estimated to be over 60 times higher than in high-income countries [1,2], reflecting disparities in risk factors, healthcare access, and disease management [3–5]. The aetiology of childhood pneumonia is multifactorial [6,7], with both viral and bacterial pathogens contributing to the burden of disease [7,8]. Respiratory syncytial virus (RSV) and influenza viruses are among the most frequently identified viral causes [7,9], while Streptococcus pneumoniae and Haemophilus influenzae, particularly non-vaccine types, are the most important bacterial causes of pneumonia [7,10].

The upper respiratory tract (URT) serves as the first point of contact between the host and the pathogens that cause pneumonia [11]. It is a complex ecosystem comprising both commensal and potentially pathogenic bacteria and plays a central role in respiratory health and disease by acting as a reservoir from which the lower respiratory tract may be seeded with potentially pathogenic bacteria [11–14]. Studies in high-income settings have associated respiratory illness with increased abundance of pathobionts, such as Haemophilus influenzae and Streptococcus pneumoniae [11,15–17], and reduced levels of protective taxa such as Dolosigranulum pigrum, Corynebacterium spp., and Moraxella lincolnii [18–20]. However, these insights are derived almost exclusively from children in high-income countries, where demographic and social characteristics as well as environmental exposures differ substantially from children in LMICs, particularly sub-Saharan Africa [11].

In many African countries, additional risk factors such as severe malnutrition, HIV exposure, indoor air pollution, and household structure and size may exert profound effects on the respiratory microbiota and modify the risk of developing severe pneumonia [11,21]. To date, very few studies have analysed the association between the respiratory microbiome and these clinical and demographic factors within this setting. To address this gap, we conducted a comprehensive, multicentre, cross-sectional analysis of the nasopharyngeal microbiome in children hospitalized with severe pneumonia in East Africa. This study was nested within the Children’s Oxygen Administration Strategies Trial (COAST), a large randomized controlled trial evaluating oxygen delivery strategies in children with pneumonia in six hospitals in Kenya and Uganda [22]. Within this trial framework, we analysed a subset of children whose upper airways were sampled by nasopharyngeal swabbing at admission and performed bacterial community profiling using partial 16S rRNA gene sequencing. Clinical and demographic data were collected prospectively, and multiplex PCR was performed for a standardized panel of respiratory viruses. In this descriptive analysis, we sought to characterize the composition and structure of nasopharyngeal microbiota in this population and explore its associations with key clinical features, including HIV status, bacteraemia, malaria, sickle cell disease, sepsis, respiratory viral infections (RSV, Influenza and hMPV), and mortality. We provide among the first high-resolution map of the paediatric respiratory microbiome in a large-scale LMIC pneumonia cohort.

Methods

Study design, setting, and participants

This cross-sectional respiratory microbiome profiling study was nested within the Children’s Oxygen Administration Strategies Trial (COAST; Clinical trial No: ISRCTN15622505), a multicentre randomized controlled trial that evaluated different oxygen administration strategies in children with severe pneumonia [22]. The COAST trial [22] was conducted in six hospitals in East Africa: two in Kenya (Kilifi County Hospital and Coast General Teaching and Referral Hospital) and four in Uganda (Mulago National Referral Hospital, Jinja Regional Referral Hospital, Mbale Regional Referral Hospital, and Soroti Regional Referral Hospital). A subset of children aged between 28 days and 12 years who were admitted with life-threatening pneumonia and had nasopharyngeal (NP) samples collected at the point of hospital admission were selected for microbial profiling by partial 16S rRNA gene sequencing. Ethical approval was obtained from the Scientific and Ethics Review Unit of the Kenya Medical Research Institute (KEMRI) and the Uganda National Council for Science and Technology (UNCST). Written informed consent was obtained from parents and guardians in accordance with GCLP guidelines. NP samples were stored at −80°C immediately after collection for later analysis. Alongside NP sampling, detailed clinical and demographic metadata were captured, including nutritional status, vital signs, malaria status, HIV status, and outcomes such as discharge from hospital or in-hospital death. Two radiologists independently read the chest radiographs (CXRs), and a third reader resolved discrepancies and classified for presence of consolidation. A combined approach was used to define sepsis, involving diagnosis by the admitting clinician and confirmed by laboratory data of abnormal white cell counts (total WBC >12 000 or <4000 cells/µL) [23]. Sickle Cell Disease (SCD) status was confirmed by molecular genotyping. In addition, multiplex PCR testing was performed for 15 respiratory pathogens, including RSV A and B, influenza A, B, and C, parainfluenza viruses 1–4, human metapneumovirus, rhinovirus, adenovirus, coronavirus OC43 and 229E, and Mycoplasma pneumoniae.

DNA extraction and 16S rRNA gene sequencing

Frozen NP swabs were thawed on ice and vortexed on a Vortex-Genie 2 mixer (Labgene Scientific). DNA was extracted using the QIAamp DNA Mini Kit (Qiagen), following mechanical lysis with glass beads on the TissueLyser II (Qiagen) and enzymatic digestion with proteinase K. DNA was eluted in AE buffer and quantified using a Qubit 4 Fluorometer (Invitrogen). Amplification of the V3–V4 hypervariable region of the bacterial 16S rRNA gene was performed using 341F and 785R primers in a 12.5 µL PCR reaction. PCR products were visualized on a 2% agarose gel, purified with AMPure XP beads (Beckman Coulter), and indexed using the Nextera XT Index Kit v2 (Illumina). Indexed libraries were purified, quantified using the Qubit dsDNA HS Assay Kit, and their size distribution assessed on the Agilent 4150 TapeStation. Libraries were normalized, pooled, and denatured with NaOH, spiked with 5% PhiX control, and sequenced on the Illumina MiSeq platform (2×300 bp, MiSeq v3 kit). Two negative controls (at extraction and PCR stages) were included for quality control.

Data availability

The 16S rRNA sequence data generated from MiSeq sequencing in this study is available at NCBI SRA under accession number PRJNA1285267.

Data analysis

All analyses were performed using R (v4.5.0). Raw paired-end reads were processed using the DADA2 pipeline (v1.22.0). Quality filtering and adapter trimming were performed using filterAndTriminterface of the fastqPairedFilter function, followed by error model learning with learnErrors and denoising via dada function. Forward and reverse reads were merged using mergePairs, and chimeras were removed using removeBimeraDenovo functions. Taxonomic assignment of amplicon sequence variants (ASVs) was performed using the assignTaxonomy function and the SILVA reference database (v138). The resulting ASV table was aggregated to genus level, and samples were normalized to relative abundance. Microbiome composition was assessed using hierarchical clustering of Bray-Curtis dissimilarities computed from genus-level relative abundance profiles. Clustering was performed using Ward’s method and visualised using dendrograms and stacked bar plots of the top 50 genera. Samples were assigned to six compositional clusters, identified using the NbClust R package based on the community microbiome structure. To identify dominant taxa within each cluster, average genus abundance per cluster was computed, and genus names were rendered as bubble plots, scaled by mean relative abundance. Compositional data transformation was performed by applying centered log-ratio (CLR) on the genus features prior to Maaslin2 and temporal models analyses. This controls for the spurious correlations arising from relative abundance constraints. To explore overall microbial structure in relation to clinical and demographic metadata, two-dimensional non-metric multidimensional scaling (NMDS) was applied to Bray-Curtis dissimilarities. The envfit function from the vegan package was used to fit clinical and virological covariates as vectors into NMDS space with 999 permutations to assess significance. Significant genus vectors were overlaid on the NMDS ordination to examine alignment of different taxa and microbial clusters with ordination gradients. Associations between individual microbial taxa and clinical or virological variable were also tested using the Maaslin2 R package (v1.21.0). Relative abundances at the genus level were used as outcomes, and different clinical features (sepsis, cyanosis, death, coma, severe malaria, HIV status, SCD, asthma history, diarrhoea, malnutrition, influenza A, RSV) were passed as fixed effects in multivariable models, adjusting for age and study site as potential confounders. Significant associations were defined using an FDR-adjusted q-value < 0.25. Results were visualised using coefficient plots (with standard errors) and annotated boxplots showing genus-level relative abundance stratified by metadata variable.

Results

Study population and clinical characteristics

We analysed nasopharyngeal swab (NP) samples from 876 children aged 28 days to 12 years who were admitted with life-threatening pneumonia to five hospitals in Kenya and Uganda participating in the COAST trial. The sites included Kilifi, Mulago, Jinja, Mbale, and Soroti. Baseline characteristics varied by site, with significant differences in age distribution, breastfeeding status, nutritional indicators, and mortality. Median age ranged from 7.5 months in Soroti to 14 months in Jinja. Overall mortality was highest in Kilifi and Mulago (23.1% each), and lowest in Soroti (1.7%). Severe malaria, sepsis, and coma were more frequently recorded in children from Mbale and Kilifi compared to other sites. Baseline clinical and demographic characteristics are summarized in Table 1.

Table 1. Baseline characteristics of the study participants.

Jinja (N=78) Kilifi (N=13) Mbale (N=348) Mulago (N=13) Soroti (N=424) P value
Demographic characteristics
    Age in months, median (IQR) 14 (7-38.2) 8 (3-15) 8 (4-22) 8 (3-15) 7.5 (3-18) 0.002
    Female, n (%) 36 (46.2) 4 (30.8) 150 (43.1) 2 (15.4) 176 (41.5) 0.265
    Breasted, n(%) 78 (100%) 12 (92.3) 347 (99.7) 13 (100) 343 (80.9) <0.001
    Siblings, median (IQR) 2 (1-3.8) 4 (2-5) 2 (1-3) 1 (0-2) 2 (1-4) <0.001
Clinical characteristics
    Temperature, median (IQR) 38 (37–38.5) 37 (37–38) 38 (37–38) 37 (36–37) 37 (37–38) 0.006
    Heart rate, median (IQR) 167.5 (143.8–182) 164(158–171) 164 (148–174) 142 (132–158) 157 (143–159) <0.001
    Respiratory rate, median (IQR) 60 (48–72) 62 (58–64) 59.5 (50.8–66) 62 (58–66) 63 (54–73) 0.005
    Oxygen saturation, median (IQR) 86 (78.2–88) 89 (86–89) 87 (81.8–89) 83 (81–88) 87 (83–89) 0.105
    Weight, median (IQR) 9 (7–14) 5 (4–9) 8 (6–10) 8 (7–11) 8 (6–10) 0.048
    Comatose, n (%) 3 (3.8) 1 (7.7) 1 (0.3) 1 (7.7) 2 (0.5) <0.001
    Diarrhoea, n (%) 13 (16.7) 3 (23.1) 69 (19.8) 0 31 (7.3) <0.001
    Severe malaria, n (%) 21 (26.9) 6 (46.2) 87 (25) 1 (7.7) 24 (5.7) <0.001
    Sepsis, n (%) 16 (20.5) 2 (15.4) 46 (13.2) 0 11 (2.6) <0.001
    Cyanosis, n (%) 2 (2.6) 0 11 (3.2) 0 2 (0.5) 0.051
Comorbidities
    Asthma, n (%) 0 1 (7.7) 3 (0.9) 0 12 (2.8) 0.079
    HIV+, n (%) 0 0 2 (0.6) 0 2 (0.5) 0.865
    Sickle cell crisis, n (%) 11 (14.1) 1 (7.7) 15 (4.3) 0 18 (4.2) 0.005
    Malnutrition, muac<11.5cm, n (%) 11 (14.1) 5 (38.5) 34 (9.8) 0 27 (6.4) <0.001
    Anaemia, n (%) 45 (95.7) 4 (30) 313 (92.9) 12 (100) 386 (93.5) 0.58
Outcome
    Died, n (%) 5 (6.4) 3 (23.1) 18 (5.2) 3 (23.1) 7 (1.7) <0.001
Open in a new tab

Clinical associations with airway microbiome structure

To identify associations between the overall structure of the airway microbiome and clinical factors, we first conducted hierarchical clustering of genus-level relative abundance profiles and identified six distinct microbiome clusters across the 876 samples. Clustering was based on Bray-Curtis dissimilarity and Ward’s linkage method, with visualization by dendrograms and stacked bar plots of the top 50 genera as shown in Fig 1. Six distinct clusters were identified, each displaying a unique microbial composition pattern, with one or two dominant genera typically dominating specific clusters. Bubble plot visualizations (Fig 2a) showed that cluster 1 was dominated by Staphylococcus, cluster 2 by Streptococcus, while cluster 3 was heterogeneous and exhibited no clear patterns of dominance by a single taxon. Cluster 4 was dominated by Haemophilus, cluster 5 by Dolosigranulum and Corynebacterium, and cluster 6 by Moraxella (Fig 2a). To explore the relationships between this microbial structure and host factors, we applied non-metric multidimensional scaling (NMDS) to Bray-Curtis dissimilarities of genus-level profiles. Microbiota clusters 1, 2, and 5 showed clear spatial separation in ordination space, reflecting distinct compositional patterns described above (Fig 2b). Clinical and demographic covariates were then projected into the same ordination space. Clinical outcomes including influenza A infection and death were significantly associated with ordination gradients (p < 0.05) as shown in Fig 2b. Taxonomic vectors revealed directional alignments with clinical outcomes: death was directionally associated with cluster 5 (dominated by Corynebacterium and Dolosigranulum), while influenza A infection aligned with cluster 1 (dominated by Staphylococcus) as shown in Fig 2b.

Figure 1.

Figure 1

Open in a new tab

Hierarchical clustering of genus-level relative abundance profiles from 876 children revealed six distinct nasopharyngeal microbiome clusters based on Bray–Curtis dissimilarity and Ward’s linkage. Heatmap displays the top 50 genera; columns represent individual samples and rows represent bacterial genera.

Figure 2. Airway microbiota structure in children hospitalized with severe pneumonia.

Figure 2

Open in a new tab

(a) Bubble plots showing dominant genera in each cluster, scaled by mean relative abundance within cluster. Clusters were characterized by the following dominant genera: Staphylococcus (Cluster 1), Streptococcus (Cluster 2), Haemophilus (Cluster 4), Corynebacterium and Dolosigranulum (Cluster 5), and Moraxella (Cluster 6). Cluster 3 lacked a dominant genus and exhibited heterogeneous composition.

(b) Non-metric multidimensional scaling (NMDS) of Bray–Curtis dissimilarities overlaid with microbial cluster assignments and genus vectors. Arrows represent clinical outcomes significantly associated with gradients in microbiota structure (envfit p < 0.05), including death and influenza A infection. Death vector aligned directionally with Cluster 5, while influenza A infection aligned with Cluster 1. Age and study site were significantly associated with the ordination space, indicating their potential to confound associations with clinical outcomes such as mortality and viral infection.

Association between bacterial genera and clinical outcomes

Microbiome data are compositional, with the taxa abundancies expressed as relative proportions constrained to a fixed total, potentially creating false associations as changes in one taxon could reflect a shift in another rather than true biological differences. We performed centered log-ratio transformation to reduce this bias and improve interpretability, though results should still be considered within compositional data limitations. To identify associations between specific genera and clinical outcomes, we performed multivariable regression modelling using Maaslin2 after adjusting for age and study site [24]. Relative abundance of genera was modelled as the outcome, with fixed effects including sepsis, cyanosis, death, coma, severe malaria, HIV status, SCD, asthma history, diarrhoea, malnutrition, RSV, and influenza A infection. Significant associations (p-value < 0.05, FDR q < 0.25) are shown in Fig 3a. Corynebacterium showed a positive beta association in children who died, indicating an increased abundance of this genus in the airways of children with this outcome. Streptococcus, on the other hand, was negatively associated with influenza A (Fig 3a). Similarly, Dolosigranulum was negatively associated with Influenza A infections, reflecting lower bacterial abundance in infected children, while Gemella showed a positive beta association among HIV positive children, indicating higher abundance in these children relative to the rest of the study population. Haemophilus was negatively associated with culture-confirmed bacteraemia These associations are shown using boxplots stratified by clinical group in Fig 3b.

Figure 3. Associations between specific bacterial genera and clinical outcomes.

Figure 3

Open in a new tab

(a) Results from multivariable Maaslin2 models showing associations between genus-level relative abundance and clinical or virological covariates. Beta coefficients and standard errors are shown for selected genera with a false discovery rate (FDR)–adjusted q-value < 0.25. Corynebacterium abundance was positively associated with death; Gemella was elevated in children living with HIV. Streptococcus was reduced in children infected with influenza A and Haemophilus was similarly reduced in children with culture-confirmed bacteraemia.

(b) Annotated boxplots showing genus-level relative abundance stratified by clinical status for selected significant associations from panel (a).

Alignment of microbial composition with time to death

To further explore the association between airway Corynebacterium abundance and death, we stratified children who died into five distinct time strata based on the duration of time from admission to death as follows: children who died on the day of admission, children who died one or two days after admission, children who died between 3 and 7 days after admission, and children who died between 8 and 35 days after admission. We then used nonlinear regression modelling using loess fitting to visualise changes in microbial abundance relative to the time of death. Of all the genera analysed, only Corynebacterium showed a clear temporal trend relative to the time of death: the highest abundance was observed among those who died on the day of admission, dropping progressively among children with longer survival times. Median Corynebacteria abundance in children who died ≥ 3 days after admission was similar to that of survivors (Fig 4a). However, this temporal trend was not statistically significant in the generalized additive mixed models treating time to death as a continuous variable and adjusting for age and study site (Fig 4b), suggesting that the observed elevated corynebacterium abundance in early deaths is driven by subgroup effects and not a continuous trend over time to death intervals.

Figure 4. Temporal patterns of genus-level abundance by time to death.

Figure 4

Open in a new tab

(a) LOESS smoothed trajectories of genus-level relative abundance in children who died, stratified by time from admission to death, compared to median abundance values in survivors (red dashed lines). Corynebacterium abundance was highest in early deaths (day 0–2), followed by a progressive decline among later deaths, converging toward survivor levels. Dolosigranulum, Haemophilus, and Streptococcus abundances showed no consistent temporal variation relative to time of death. Shaded ribbons denote 95% confidence intervals for loess-smoothed estimates.

(b) Results from generalized additive mixed models (GAMMs) showing the relationship between time to death (days) from admission and genus-level relative abundance for Corynebacterium, Dolosigranulum, Haemophilus, and Streptococcus. Time to death was treated as a continuous variable and included age as a fixed effect and study site as a random effect. The blue lines denote the fitted GAMM curves and the grey shaded region show 95% confidence level. Dolosigranulum and Haemophilus show a declining trend with proximity to death, although not significant.

Discussion

This study presents one of the largest high-resolution surveys of the paediatrics nasopharyngeal (NP) microbiome in sub-Saharan Africa, leveraging data from 876 children hospitalized with life-threatening pneumonia in five hospital sites in Kenya and Uganda. Using partial 16S rRNA gene sequencing, clinical phenotyping, and virological testing, we identified six microbiome clusters, characterized by varying taxonomic composition, and explored their relationships with key clinical outcomes, including sepsis, cyanosis, death, coma, severe malaria, HIV status, SCD, asthma history, diarrhoea, malnutrition, respiratory viral infections and death. This analysis showed significant associations between the airway microbiome structure and clinical outcomes, including death, HIV status, bacteraemia and influenza A infection. Genus-level associations resolved using multivariable regression modelling, adjusting for study site and age showed outcomes such as death and HIV status were significantly associated with different microbial taxa, including Corynebacterium and Gemella, respectively.

An unexpected finding in our study was the association between Corynebacterium abundance and increased mortality, particularly in children who died early (< 3 days) during hospitalization. LOESS-smoothed temporal analysis showed striking early elevation in Corynebacterium abundance in children who died within 48 hours, followed by a progressive decline to levels that were comparable to those of survivors among children who died later. However, this observation was not statistically significant in the generalized additive mixed model by treating time to death as a continuous variable. This suggests the cross-sectional signal is driven by a local effect that is specific to early death and not a sustained trend across all death timepoints. This contrasts with the presumed beneficial role of this genus in HIC cohorts and experimental studies [25–29] and a recent study of South African children where Corynebacterium was associated with reduced risk of lower respiratory infection [30]. This may reflect differences in species level composition, underlying host factors, and geographic exposures. One speculative explanation is that the species composition of Corynebacterium differs substantially in LMICs. For example, Corynebacterium species such as Corynebacterium ulcerans, a known zoonotic pathogen [31], is more prevalent in rural African settings [32,33] due to cultural nuances that enhance human-animal interaction. C. ulcerans has been implicated in severe respiratory infections [34], including pneumonia [35], and produces diphtheria toxin in some cases [31,36,37]. It is typically acquired through close contact with livestock or the consumption of unpasteurized milk [37–39] - both common in rural African households where cohabitation with animals and use of raw milk as a nutritional supplement are widespread practices [40]. Recent case series from Japan and Europe have reported increasing incidence and mortality associated with C. ulcerans respiratory infections [35,39], underscoring its potential clinical impact. Given that our cohort included children from settings where such exposures are common, it is plausible that elevated Corynebacterium abundance may reflect carriage or infection with pathogenic species such as C. ulcerans. Unfortunately, partial 16S rRNA gene sequencing is not sufficiently sensitive for species-level resolution [41], and this hypothesis will therefore require confirmation using higher-resolution shotgun metagenomics approaches or other targeted molecular methods. We also stratified non-survivors by time to death and compared genus-specific abundance profiles among children with different survival times to those of survivors. The airway abundance of Dolosigranulum, Haemophilus and Streptococcus, on the other hand, did not exhibit a similar temporal pattern relative to the time of death and remained comparable to corresponding median survivor abundances at all timepoints. This temporal pattern suggests that elevated Corynebacterium abundance at admission may be a marker of acute disease severity, or a dysbiotic microbial state associated with imminent deterioration. In contrast, the absence of clear temporal variation in Dolosigranulum, Haemophilus, and Streptococcus abundance suggests that their associations with mortality are either weak, non-specific, or masked by clinical heterogeneity. Together, these findings raise the possibility that the abundance of certain Corynebacterium species may serve as a context-sensitive indicator of early, fulminant disease trajectories in this setting, although species-specific attributions will require confirmation using shotgun approaches as discussed above.

Several limitations of this study warrant consideration. First, this was a cross-sectional analysis at hospital admission, and therefore causality could not be inferred. Second, partial 16S rRNA sequencing does not resolve species or functional potential, limiting taxonomic specificity and precluding inference of virulence mechanisms. For instance, Haemophilus abundance was negatively associated with bacteraemia. While counterintuitive given the established role of Haemophilus influenzae as an invasive respiratory pathogen, our 16S partial sequencing approach cannot distinguish pathogenic H. influenzae from commensal Haemophilus species. Importantly, all the nasopharyngeal samples in this study are archived in our biobank and available for future shotgun metagenomics for strain-level resolution and functional profiling. Third, our data are drawn from hospitalized children with severe disease and may not generalize to mild community-acquired pneumonia or healthy controls. Nonetheless, this study provides one of the most comprehensive descriptions of the NP microbiome in African children with pneumonia to date. By combining deep clinical phenotyping with community profiling across diverse settings, we reveal both shared and context-specific features of microbial ecology in paediatric pneumonia. These findings set the stage for future mechanistic studies and highlight the importance of geographic and ecological diversity in microbiome research.

Article Summary.

This study investigates the nasal microbiota in young children in East Africa with severe pneumonia revealing the bacteria linked to severe disease outcomes in low resource settings

Acknowledgments

We thank all the study participants and staff from all the centres participating in the COAST trial. We thank the Trial Steering Committee for its support of the trial.

Funding

This study is part of the EDCTP2 programme (grant number RIA-2016S-1636-COAST-Nutrition) supported by the European Union and also supported by a grant (MR/R018502/1) from the Joint Global Health Trials scheme, which included the U.K. Medical Research Council, the U.K. Foreign, Commonwealth and Development Office, NIHR using UK international development funding from the UK Government and the Wellcome Trust.

Footnotes

Author contribution statement

Timothy O Makori: Conceptualization and design, methodology, laboratory and bioinformatics data analysis, visualization, drafting and review of the article

Sarah Kiguli, Peter Olupot-Olupot, Rita Muhindo: Field data collection and review of the article

Christabel Mogaka: Data management (data curation) and review of the article

Martin Mutunga, Clement Lewa & Robinson Cheruiyot: Laboratory analysis (viral diagnostics)

Elijah T. Gicheru, Maureen W. Mburu, Mercy S. Sada, Omar Nyawa: Laboratory analysis (partial 16S rRNA sequencing) and article review

Thomas N Williams, Charles N. Agoti: Funding acquisition, conceptualization and design, data acquisition, and review of the article

Kathryn Maitland: Funding acquisition, conceptualization and design, data acquisition, supervision, review of the article

Charles N Sande: Funding acquisition, conceptualization and design, data acquisition, supervision, drafting and review of article

Conflict of interest: None

Data Availability

The data supporting the findings of this study are available from Harvard Dataverse (https://doi.org/10.7910/DVN/L2C5E6).

References

  • 1.McAllister DA, Liu L, Shi T, et al. Global, regional, and national estimates of pneumonia morbidity and mortality in children younger than 5 years between 2000 and 2015: a systematic analysis. Lancet Glob Health. 2019 Jan;7(1):e47–57. doi: 10.1016/S2214-109X(18)30408-X. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Marangu D, Zar HJ. Childhood pneumonia in low-and-middle-income countries: An update. Paediatr Respir Rev. 2019 Nov;32:3–9. doi: 10.1016/j.prrv.2019.06.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Ahmed S, Mvalo T, Akech S, et al. Protecting children in low-income and middle-income countries from COVID-19. [cited 2025 Jun 10];BMJ Glob Health. 2020 May 26;5(5):50. doi: 10.1136/bmjgh-2020-002844. [Internet] Available from: https://gh.bmj.com/content/5/5/e002844. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Simkovich SM, Underhill LJ, Kirby MA, et al. Resources and Geographic Access to Care for Severe Pediatric Pneumonia in Four Resource-limited Settings. [cited 2025 Jun 10];Am J Respir Crit Care Med. 2022 Jan 15;205(2):183–97. doi: 10.1164/rccm.202104-1013OC. [Internet] Available from. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Nemet M, Gmehlin C, Vukoja M, Dong Y, Gajic O, Tekin A, Checklist for Early Recognition and Treatment of Acute Illness and Injury (CERTAIN) study investigators Ventilator-Associated Pneumonia in Low- and Middle-Income vs. High-Income Countries: The Role of Ventilator Bundle, Ventilation Practices, and Healthcare Staffing. [cited 2025 Jun 10];Chest. 2025 doi: 10.1016/j.chest.2025.01.002. [Internet] Available from. [DOI] [PubMed] [Google Scholar]
  • 6.Yun KW. Community-acquired pneumonia in children: updated perspectives on its etiology, diagnosis, and treatment. Clin Exp Pediatr. 2024 Feb 15;67(2):80–9. doi: 10.3345/cep.2022.01452. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Torres A, Cilloniz C, Niederman MS, Menéndez R, Chalmers JD, Wunderink RG, van der Poll T. Pneumonia. Nat Rev Dis Primers. 2021 Apr 8;7(1):25. doi: 10.1038/s41572-021-00259-0. [DOI] [PubMed] [Google Scholar]
  • 8.Yadav KK, Awasthi S. Childhood Pneumonia: What’s Unchanged, and What’s New? [cited 2025 Jun 10];Indian J Pediatr. 2023 Jul 1;90(7):693–9. doi: 10.1007/s12098-023-04628-3. [Internet] Available from. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.de Benedictis FM, Kerem E, Chang AB, Colin AA, Zar HJ, Bush A. Complicated pneumonia in children. The Lancet. 2020 Sep;396(10253):786–98. doi: 10.1016/S0140-6736(20)31550-6. [DOI] [PubMed] [Google Scholar]
  • 10.Cilloniz C, Martin-Loeches I, Garcia-Vidal C, Jose AS, Torres A. Microbial Etiology of Pneumonia: Epidemiology, Diagnosis and Resistance Patterns. [cited 2025 Jun 10];International Journal of Molecular Sciences. 2016 Dec 16;17(12):2120. doi: 10.3390/ijms17122120. 2016, Vol 17, Page 2120 [Internet] Available from: https://www.mdpi.com/1422-0067/17/12/2120/htm. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Man WH, de Steenhuijsen Piters WAA, Bogaert D. The microbiota of the respiratory tract: gatekeeper to respiratory health. Nat Rev Microbiol. 2017 May 20;15(5):259–70. doi: 10.1038/nrmicro.2017.14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Kumpitsch C, Koskinen K, Schöpf V, Moissl-Eichinger C. The microbiome of the upper respiratory tract in health and disease. BMC Biol. 2019 Dec 7;17(1):87. doi: 10.1186/s12915-019-0703-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Santacroce L, Charitos IA, Ballini A, Inchingolo F, Luperto P, De Nitto E, Topi S. The Human Respiratory System and its Microbiome at a Glimpse. Biology (Basel) 2020 Oct 1;9(10):318. doi: 10.3390/biology9100318. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Claassen-Weitz S, Lim KYL, Mullally C, Zar HJ, Nicol MP. The association between bacteria colonizing the upper respiratory tract and lower respiratory tract infection in young children: a systematic review and meta-analysis. Clinical Microbiology and Infection. 2021 Sep;27(9):1262–70. doi: 10.1016/j.cmi.2021.05.034. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Seppanen EJ, Bayliss J, Clark SL, et al. Haemophilus influenzae remains the predominant otitis media pathogen in Australian children undergoing ventilation tube insertion in the PCV13 era. [cited 2025 Jun 12];Journal of Infection. 2025 Apr 1;90(4):106478. doi: 10.1016/j.jinf.2025.106478. [Internet] Available from: https://www.sciencedirect.com/science/article/pii/S0163445325000726. [DOI] [PubMed] [Google Scholar]
  • 16.Bogaert D, De Groot R, Hermans PWM. Streptococcus pneumoniae colonisation: the key to pneumococcal disease. [cited 2025 Jun 12];Lancet Infect Dis. 2004 Mar 1;4(3):144–54. doi: 10.1016/S1473-3099(04)00938-7. [Internet] Available from: https://www.thelancet.com/action/showFullText?pii=S1473309904009387. [DOI] [PubMed] [Google Scholar]
  • 17.Laufer AS, Metlay JP, Gent JF, Fennie KP, Kong Y, Pettigrewa MM. Microbial communities of the upper respiratory tract and otitis media in children. [cited 2025 Jun 12];mBio. 2011 2(1):245–55. doi: 10.1128/mBio.00245-10. [Internet] Available from. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.de Steenhuijsen Piters WAA, Binkowska J, Bogaert D. Early Life Microbiota and Respiratory Tract Infections. Cell Host Microbe. 2020 Aug;28(2):223–32. doi: 10.1016/j.chom.2020.07.004. [DOI] [PubMed] [Google Scholar]
  • 19.Henares D, Brotons P, de Sevilla MF, et al. Differential nasopharyngeal microbiota composition in children according to respiratory health status. Microb Genom. 2021 Nov 1;7(10) doi: 10.1099/mgen.0.000661. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Lapidot R, Faits T, Ismail A, et al. Nasopharyngeal Dysbiosis Precedes the Development of Lower Respiratory Tract Infections in Young Infants, a Longitudinal Infant Cohort Study. Gates Open Res. 2024 Mar 20;6:48. doi: 10.12688/gatesopenres.13561.2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Dubowski K, Kaali S, Jack D, et al. Infant nasopharyngeal microbiota subphenotypes and early childhood lung function: Evidence from a rural ghanaian pregnancy cohort. [cited 2025 Jun 12];Int J Environ Res Public Health. 2021 Jul 2;18(14):7276. doi: 10.3390/ijerph18147276. [Internet] Available from: https://www.mdpi.com/1660-4601/18/14/7276/htm. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Maitland K, Kiguli S, Opoka RO, et al. Children’s Oxygen Administration Strategies Trial (COAST): A randomised controlled trial of high flow versus oxygen versus control in African children with severe pneumonia. [cited 2025 Jun 12];Wellcome Open Res. 2018 2:100. doi: 10.12688/wellcomeopenres.12747.2. [Internet] Available from. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Singer M, Deutschman CS, Seymour C, et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3) [cited 2025 Aug 22];JAMA. 2016 Feb 23;315(8):801–10. doi: 10.1001/jama.2016.0287. [Internet] Available from: https://jamanetwork.com/journals/jama/fullarticle/2492881. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Mallick H, Rahnavard A, McIver LJ, et al. Multivariable association discovery in population-scale meta-omics studies. [cited 2025 Jun 17];PLoS Comput Biol. 2021 Nov 1;17(11):e1009442. doi: 10.1371/journal.pcbi.1009442. [Internet] Available from. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Bosch AATM, Levin E, van Houten MA, et al. Development of Upper Respiratory Tract Microbiota in Infancy is Affected by Mode of Delivery. [cited 2025 Jun 17];EBioMedicine. 2016 Jul 1;9:336–45. doi: 10.1016/j.ebiom.2016.05.031. [Internet] Available from: https://www.sciencedirect.com/science/article/pii/S2352396416302250#s0040. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Biesbroek G, Tsivtsivadze E, Sanders EAM, Montijn R, Veenhoven RH, Keijser BJF, Bogaert D. Early respiratory microbiota composition determines bacterial succession patterns and respiratory health in children. [cited 2025 Jun 17];Am J Respir Crit Care Med. 2014 Dec 1;190(11):1283–92. doi: 10.1164/rccm.201407-1240OC. [Internet] Available from. [DOI] [PubMed] [Google Scholar]
  • 27.Kanmani P, Clua P, Vizoso-Pinto MG, et al. Respiratory commensal bacteria Corynebacterium pseudodiphtheriticum improves resistance of infant mice to respiratory syncytial virus and Streptococcus pneumoniae superinfection. [cited 2025 Jun 17];Front Microbiol. 2017 Aug 23;8:273272. doi: 10.3389/fmicb.2017.01613. [Internet] Available from: www.frontiersin.org. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Horn KJ, Jaberi Vivar AC, Arenas V, Andani S, Janoff EN, Clark SE. Corynebacterium Species Inhibit Streptococcus pneumoniae Colonization and Infection of the Mouse Airway. [cited 2025 Jun 17];Front Microbiol. 2022 Jan 10;12:804935. doi: 10.3389/fmicb.2021.804935. [Internet] Available from: www.frontiersin.org. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Bomar L, Brugger SD, Yost BH, Davies SS, Lemon KP. Corynebacterium accolens releases antipneumococcal free fatty acids from human nostril and skin surface triacylglycerols. [cited 2025 Jun 17];mBio. 2016 Jan 5;7(1) doi: 10.1128/mBio.01725-15. [Internet] Available from. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Claassen-Weitz S, Xia Y, Hannan L, et al. Nasopharyngeal Microbiota in South African Infants With Lower Respiratory Tract Infection: A Nested Case-Control Study of the Drakenstein Child Health Study. [cited 2025 Dec 16];Clinical Infectious Diseases. 2025 doi: 10.1093/cid/ciaf184. [Internet] Available from. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Crestani C, Passet V, Rethoret-Pasty M, et al. Microevolution and genomic epidemiology of the diphtheria-causing zoonotic pathogen Corynebacterium ulcerans. [cited 2025 Jun 18];Nature Communications. 2025 May 24;16(1):1–11. doi: 10.1038/s41467-025-60065-0. 2025 16:1 [Internet] Available from: https://www.nature.com/articles/s41467-025-60065-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Addo PB, Dennis SM. Corynebacteria Associated with Diseases of Cattle, Sheep and Goats in Northern Nigeria. [cited 2025 Jun 18];British Veterinary Journal. 1977 Jul 1;133(4):334–9. doi: 10.1016/s0007-1935(17)34032-0. [Internet] Available from: https://www.sciencedirect.com/science/article/abs/pii/S0007193517340320. [DOI] [PubMed] [Google Scholar]
  • 33.World Health Organization. Disease Outbreak News; Diphtheria in Nigeria. 2023. Sep, [cited 2025 Jun 18]. [Internet] Available from: https://www.who.int/emergencies/disease-outbreak-news/item/2023-DON485.
  • 34.Otshudiema JO, Acosta AM, Cassiday PK, Hadler SC, Hariri S, Tiwari TSP. Respiratory Illness Caused by Corynebacterium diphtheriae and C. ulcerans, and Use of Diphtheria Antitoxin in the United States, 1996-2018. [cited 2025 Jun 18];Clinical Infectious Diseases. 2021 Nov 1;73(9):E2799–806. doi: 10.1093/cid/ciaa1218. [Internet] Available from. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Yasuda I, Matsuyama H, Ishifuji T, et al. Severe Pneumonia Caused by Toxigenic Corynebacterium ulcerans Infection, Japan. [cited 2025 Jun 18];Emerg Infect Dis. 2018 Mar 1;24(3):588. doi: 10.3201/eid2403.171837. [Internet] Available from. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Sekizuka T, Yamamoto A, Komiya T, et al. Corynebacterium ulcerans 0102 carries the gene encoding diphtheria toxin on a prophage different from the C. diphtheriae NCTC 13129 prophage. [cited 2025 Jun 18];BMC Microbiol. 2012 May 14;12(1):1–7. doi: 10.1186/1471-2180-12-72. [Internet] Available from. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Hillan A, Gibbs T, Weaire-Buchanan G, Brown T, Pang S, McEvoy SP, Parker E. Zoonotic transmission of diphtheria toxin-producing Corynebacterium ulcerans. [cited 2025 Jun 18];Zoonoses Public Health. 2024 Mar 1;71(2):157–69. doi: 10.1111/zph.13094. [Internet] Available from. [DOI] [PubMed] [Google Scholar]
  • 38.Hoefer A, Herrera-León S, Domínguez L, et al. Zoonotic Transmission of Diphtheria from Domestic Animal Reservoir, Spain. [cited 2025 Jun 18];Emerg Infect Dis. 2022 Jun 1;28(6):1257. doi: 10.3201/eid2806.211956. [Internet] Available from. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Zendri F, Isgren CM, Sinovich M, et al. Case Report: Toxigenic Corynebacterium ulcerans Diphtheria-Like Infection in a Horse in the United Kingdom. [cited 2025 Jun 18];Front Vet Sci. 2021 Jun 1;8:650238. doi: 10.3389/fvets.2021.650238. [Internet] Available from: https://www.eucast.org/ [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Mpatswenumugabo JP, Mukasafari MA, Ndahetuye JB, Wredle E, Båge R. A systematic literature review of milk consumption and associated bacterial zoonoses in East Africa. [cited 2025 Jun 18];J Appl Microbiol. 2023 Apr 3;134(4):1–13. doi: 10.1093/jambio/lxad080. [Internet] Available from. [DOI] [PubMed] [Google Scholar]
  • 41.Johnson JS, Spakowicz DJ, Hong BY, et al. Evaluation of 16S rRNA gene sequencing for species and strain-level microbiome analysis. [cited 2025 Jun 18];Nature Communications. 2019 Nov 6;10(1):1–11. doi: 10.1038/s41467-019-13036-1. 2019 10:1 [Internet] Available from: https://www.nature.com/articles/s41467-019-13036-1. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

The 16S rRNA sequence data generated from MiSeq sequencing in this study is available at NCBI SRA under accession number PRJNA1285267.

The data supporting the findings of this study are available from Harvard Dataverse (https://doi.org/10.7910/DVN/L2C5E6).

RESOURCES