BPG is committed to discovery and dissemination of knowledge
Review
©The Author(s) 2026.
World J Gastroenterol. Jan 21, 2026; 32(3): 112437
Published online Jan 21, 2026. doi: 10.3748/wjg.v32.i3.112437
Table 4 Studies investigating microbiota and coronavirus disease 2019
Ref.
COVID-related cases (n)
Methodology
Increased taxa
Decreased taxa
Limitations
Conclusion
de Nies et al[119], 2023118 subjects COVID-19: 61 (asymptomatic-moderate). Control: 57Metagenomics, metatranscriptomics, MAG reconstruction, VF/ARG prediction, virome analysisPrevotella stercorea, Prevotella spp. CAG 520, Roseburia spp. CAG 471, Firmicutes (AM10); Betaherpesvirus; Rotavirus C; VFs & AMR genes (Acidaminococcaceae, Erysipelatoclostridiaceae)Turicibacter sanguinis, Roseburia faecis, Firmicutes (CAG 145)Limited to asymptomatic-moderate cases; single-region cohort (Luxembourg); unclear mechanisms linking SARS-CoV-2 to VF/ARG expressionCOVID-19 increases the virulence and AMR potential of commensals despite minimal taxonomic shifts, suggesting enhanced pathogenic potential of the gut microbiota
Li et al[120], 2025121 participants total, NC: 53 (no COVID-19), C3M: 27 (3 months post-recovery), C6M: 41 (6 months post-recovery)Metagenomic shotgun sequencing; ITS sequencing for fungiBlautia massiliensis, Kluyveromyces spp., Bacteroides xylanisolvens, Phocaeicola vulgatus, Weissella confusa, Streptococcus thermophilus, Asterotremella spp., Gibberella spp.Blautia wexlerae, Bifidobacterium pseudocatenulatum, Bifidobacterium longum, Eubacterium rectale, Anaerobutyricum hallii, Pyrenochaeta spp.Cross-sectional design; excluded long COVID; smaller 3-month group; not generalizable to severe or unvaccinated casesMild COVID-19 induces long-term gut bacterial and fungal alterations lasting ≥ 6 months, with partial recovery and persistence of some pathogens
Zhang et al[121], 2023187 recovered patients (84 symptomatic)16S rRNA sequencing; clinical surveys (SF-36, SAS, SDS); lab tests; pulmonary function; chest CTVeillonellaSCFA-producers: Eubacterium hallii group, Subdoligranulum, Ruminococcus, Dorea, Coprococcus, Eubacterium ventriosum group, AgathobacterSingle-center, cross-sectional; no longitudinal monitoring; diet/Lifestyle not fully controlled; mechanisms not clarifiedLong COVID (approximately 45% at 1 year) is linked to persistent dysbiosis marked by depletion of SCFA-producing commensals, correlating with impaired quality of life, anxiety/depression, and immune dysregulation, supporting gut-lung and gut-brain axis involvement
Ishizaka et al[122], 202456 total, PLWH-CoV: 12 (mild: 7; moderate/severe: 5), PLWH controls: 25, HCs: 1916S rRNA sequencingAcute: Enterococcus faecium. Recovery (1-3 months): Roseburia, Lachnospiraceae_unclassified, Faecalibacterium prausnitzii, Eubacterium rectale. Recovered vs long COVID: Prevotella spp.Acute vs HC: Roseburia, Lachnospiraceae_unclassified. Long COVID vs recovered: Prevotella spp.Small sample size (n = 12 PLWH-CoV, only 2 PASC); no pre-infection baseline; diet, ART regimens, and variant effects not analyzedSARS-CoV-2 in PLWH causes persistent dysbiosis, marked by loss of SCFA-producers and enrichment of pathogens, with severity-linked delays in microbiome recovery and risk of PASC
Brīvība et al[123], 2024146 COVID-19 patients (92 hospitalized, 54 ambulatory) vs 110 HCsShotgun metagenomicsEnterococcus faecium, Bacteroides spp., Alistipes, EnterobacteriaceaeRoseburia, Faecalibacterium prausnitzii, Lachnospiraceae, Eubacterium rectale, Prevotella spp.High antibiotic use; heterogeneous sampling timing; phenotypic heterogeneity across patient groupsAcute COVID-19 shows reduced diversity with loss of butyrate producers; recovery involves their restoration, while Prevotella may protect against long COVID
Sorokina et al[124], 202339 post-COVID-19 patients before and after 14-day rehabilitation vs 48 healthy volunteersClinical questionnaires, CT; CBC, coagulation, biochemistry; serum IL-6, NSE (ECL); metabolites (GC-MS); microbiota (RT-PCR, colonoflor-16 kit)Bacteroides spp., Escherichia coli, Enterobacter spp., Staphylococcus aureus, IL-6, succinic acid, fumaric acid, 4-hydroxybenzoic acidLactobacillus spp., Bifidobacterium spp., Faecalibacterium prausnitzii, Phenylpropionic acidSmall cohort; no untreated controls; RT-PCR instead of sequencing; limited GI symptom assessment; diet confounded resultsPost-COVID-19 is marked by persistent dysbiosis (loss of SCFA-producers, enrichment of pathobionts) and sustained inflammatory/metabolic disturbances not resolved by standard rehabilitation, highlighting the need for personalized microbiome-targeted interventions
Tkacheva et al[125], 2023178 post-COVID-19 patients: Asymptomatic (A, n = 48), non-infected contacts (N, n = 46), severe (S, n = 86)16S rRNA sequencingRF39 (order), Clostridia UCG-014, Oscillospirales UCG-010, Akkermansia, Prevotellaceae (family), Lactobacillus, Romboutsia, Ruminococcus gnavus, ErysipelatoclostridiumParasutterella, Flavonifractor, Ruminococcus gnavus, Subdoligranulum, Methanobrevibacter, Lachnospiraceae UCG-010, Lachnospiraceae NK4A136, Barnesiella, Eubacterium xylanophilum, Eubacterium siraeumCross-sectional (3 months only); no acute phase data; diet/medications not controlled; 16S lacks functional resolutionNo major post-COVID microbiome differences by infection/severity at 3 months, but taxa correlated with immune, cardiovascular, and metabolic parameters, highlighting systemic associations beyond direct viral effects
Bredon et al[126], 2025200 COVID-19 patients vs 102 HCs (Morocco & France cohorts)Shotgun metagenomic sequencing, machine learning, metabolomics (tryptophan)Ruminococcus gnavus, Klebsiella pneumoniae, K. variicola, Bacteroides ovatus, Enterococcus; ↑ L-tryptophan biosynthesisFaecalibacterium prausnitzii, Roseburia spp., Bifidobacterium longum, Dysosmobacter welbionis, Coprococcus comes (SCFA-producers)Treatment heterogeneity (antibiotics) in the French cohort; ML model not transferable; causality not establishedCOVID-19 induces gut dysbiosis with depletion of SCFA-producers and enrichment of pathobionts, alongside altered tryptophan metabolism. Dysbiosis correlates with disease severity; the ML model predicted severity in the Moroccan cohort


Write to the Help Desk