Non-Cochrane · aetiology · PM36050830
Meta-analysis of caries microbiome studies can improve upon disease prediction outcomes.
Butcher Mark C, Short Bryn, Veena Chandra Lekha Ramalingam, et al. · APMIS : acta pathologica, microbiologica, et immunologica Scandinavica · 2022 · PubMed · doi:10.1111/apm.13272
Full textEvery question was extracted from the full text, figures and supplements by two independent LLM extractors, disagreements adjudicated. Rows cite where in the review the value comes from.
Questions
10
current version
GRADE rated
0
No studies
0
Included studies
16
linked below
No studies 0 · 0%Ungraded 10 · 100%Very low 0 · 0%Low 0 · 0%Moderate 0 · 0%High 0 · 0%
Ungraded means an effect was reported without a GRADE rating; it is not a gap. No studies means the review found nothing to answer the question. The six categories are a display order, not a score.
Review question
| Element | Planned (methods) | Found (results) |
|---|---|---|
| Population | Individuals sampled for oral (saliva/supragingival plaque) 16S rRNA microbiome in relation to dental caries | 2016 processable oral samples from 19 studies (n=19 studies; 2016 samples) |
| Intervention | Dental caries (carious oral microbiome) | Dental caries samples |
| Comparison | Caries-free / healthy oral microbiome | Caries-free / health samples |
| Outcomes | Alpha and beta diversity; microbial composition/differential abundance; predictive/diagnostic capability (random forest ROC-AUC); functional (KEGG) metabolic pathways | Alpha diversity higher in caries across all indices (p<0.001); beta diversity ADONIS significant (p<0.001) but clustering mainly by study/sequencing region; random forest could not reliably discriminate caries vs health on combined cohort (best AUC approximately 0.75, disease-only KO); predictive taxa incl. Selenomonas sp. HMT-146, Aggregatibacter actinomycetemcomitans, Actinomyces sp. HMT-896, Treponema sp. HMT-257; enriched KEGG pathways phosphonate/phosphinate and glyoxylate/dicarboxylate metabolism (p.adj=0.04) |
| Study designs | Publicly available 16S rRNA microbiome sequencing studies deposited in NCBI/ENA, published from 2012, with data score >=3 (data accessible and mappable to health/disease) | Cross-sectional (n=13), Longitudinal (n=4), Interventional (n=1), RCT (n=1) |
| Timing | Not specified | Not specified |
Search date: Not specified (studies published from 2012 onward included; no explicit search date reported). Databases: PubMed (sequence data retrieved from NCBI/ENA)
Questions and results
Dental caries (carious oral microbiome) vs Caries-free / healthy oral microbiome
Population: Individuals providing oral (saliva/supragingival plaque) 16S rRNA microbiome samples; 2016 samples from 19 studies (mixed children and adults)
| Outcome | Effect | Evidence | Status / GRADE |
|---|---|---|---|
Discrimination of caries vs health (random forest ROC-AUC) Microbiome-based random forest classifiers could not reliably discriminate caries from health across the combined cohort (best AUC approximately 0.75 in a disease-only KO subset; ungraded). Source: Results 'Assessment of diagnostic capability' and 'Summarizing individual predictability' paras; Figure 4; Figure 7A · same question in other reviews | AUC 0.75 | 19 studies · 2016 participants | Ungraded |
KEGG pathway associations Population: Participants in oral 16S rRNA caries microbiome studies Phosphonate and phosphinate metabolism and glyoxylate and dicarboxylate metabolism had the lowest reported adjusted P values (0.04); no pooled effect estimate was reported. Source: Recurrent metabolic pathways in study metadata; Figure 5 · same question in other reviews | no estimate extracted | 19 studies · 2016 participants | Ungraded |
Differential microbial abundance / taxa predictive of caries A group of genera/species (notably Selenomonas, Aggregatibacter, Actinomyces, Treponema) was predictive of caries, whereas Streptococcus was not (ungraded). Source: Results 'Organisms of interest in caries and health' para; Figure 6 · same question in other reviews | Mean-decrease-accuracy discriminating taxa: Capnocytophaga HMT-338, Lactobacillus panis, Treponema HMT-230, Bifidobacterium longum, Delftia acidovorans, Staphylococcus schleiferi; higher log2FC in caries: Selenomonas HMT-146, Aggregatibacter actinomycetemcomitans, Actinomyces HMT-896, Treponema HMT-257; Streptococcus not predictive | 19 studies · 2016 participants | Ungraded |
Alpha diversity (Observed, Chao1, Shannon, and Simpson) Population: Participants in oral 16S rRNA caries microbiome studies Across the combined sample set, alpha-diversity metrics were significantly higher in caries than in health, but no pooled effect estimate was reported. Source: Assessment of alpha and beta diversity; Figure 2 · same question in other reviews | Higher alpha diversity in caries than health; p<0.001 (Wilcoxon; no pooled effect size reported) | 19 studies · 2016 participants | Ungraded |
Beta diversity and community composition Population: Participants in oral 16S rRNA caries microbiome studies Beta-diversity matrices showed statistically significant ADONIS results, while clustering by study and 16S region was also observed; no pooled effect estimate was reported. Source: Assessment of alpha and beta diversity; Figure 3; Table S1 · same question in other reviews | no estimate extracted | 19 studies · 2016 participants | Ungraded |
Dental caries (carious oral microbiome) vs Caries-free / healthy oral microbiome
Population: Adults providing oral (saliva/supragingival plaque) 16S rRNA microbiome samples
| Outcome | Effect | Evidence | Status / GRADE |
|---|---|---|---|
Alpha diversity (Chao1, Observed, Shannon) Adults In adults, caries-vs-health alpha diversity differed for the Shannon index but not Chao1/Observed (ungraded). Source: Results 'Assessment of alpha and beta diversity' para; Figure S1 · same question in other reviews | In adults, there was no significant caries-vs-health difference in Chao1 or Observed diversity, whereas the Shannon index differed significantly (p < 0.001). | NR | Ungraded |
External disease-state discrimination Population: Oral microbiome samples from four selected case-control studies The KEGG Orthology classifier had an approximate AUC of 0.75 in the disease-only study test set. Source: Assessment of diagnostic capability of study metadata; Figure 4 · same question in other reviews | AUC 0.75 | 4 studies | Ungraded |
Dental caries (carious oral microbiome) vs Caries-free / healthy oral microbiome
Population: Children providing oral (saliva/supragingival plaque) 16S rRNA microbiome samples
| Outcome | Effect | Evidence | Status / GRADE |
|---|---|---|---|
Alpha diversity (Shannon) Children In children, caries-vs-health Shannon alpha diversity differed significantly (ungraded). Source: Results 'Assessment of alpha and beta diversity' para; Figure S1 · same question in other reviews | In children, the Shannon alpha-diversity index differed significantly between caries and health (p < 0.001). | NR | Ungraded |
Study-specific disease-state discrimination Population: Oral microbiome samples in 15 case-controlled studies Bong-Soo 2017, Zhu 2018, and Jiang 2019 had mean AUC scores over 0.75; exact study-specific AUCs were not reported in text or readable figure labels. Source: Summarizing the individual predictability of cohort studies; Figure 7 · same question in other reviews | no estimate extracted | 15 studies | Ungraded |
Comparison
| Outcome | Effect | Evidence | Status / GRADE |
|---|---|---|---|
Alpha diversity (study-level, caries vs health) Per-study across 15 studies (log2 fold change; Figure 7B) | Chao1 and Observed significantly increased in caries in only 3 of 15 studies (and pooled); overall little significance across studies | 15 studies | Ungraded |
Included primary studies 16
Open as a trial list| Study | Also included by | Outcomes it informs here | Trial PICO | Results |
|---|---|---|---|---|
| Bong-Soo, 2017 matched by author and year (probable) | this review only | not stated | not extracted yet | not extracted yet |
| Cherkasov, 2018 matched by author and year (probable) | this review only | not stated | not extracted yet | not extracted yet |
| De Jesus, 2020 matched by author and year (probable) | this review only | not stated | not extracted yet | not extracted yet |
| deJesus, 2021 matched by author and year (probable) | this review only | not stated | not extracted yet | not extracted yet |
| Ferrer, 2020 matched by author and year (probable) | this review only | not stated | not extracted yet | not extracted yet |
| Gomez, 2017 matched by author and year (probable) | this review only | not stated | not extracted yet | not extracted yet |
| Havsed, 2021 matched by author and year (probable) | this review only | not stated | not extracted yet | not extracted yet |
| Jiang, 2019 matched by author and year (probable) | this review only | not stated | not extracted yet | not extracted yet |
| Kalpana, 2020 matched by author and year (probable) | this review only | not stated | not extracted yet | not extracted yet |
| Kashimoglu, 2020 matched by author and year (probable) | this review only | not stated | not extracted yet | not extracted yet |
| Onyango, 2020 matched by author and year (probable) | this review only | not stated | not extracted yet | not extracted yet |
| Premaraj, 2020 matched by author and year (probable) | this review only | not stated | not extracted yet | not extracted yet |
| Schulz-Weidner, 2021 matched by author and year (probable) | this review only | not stated | not extracted yet | not extracted yet |
| Simon-Soro, 2018 matched by author and year (probable) | this review only | not stated | not extracted yet | not extracted yet |
| Xiao, 2016 matched by author and year (probable) | this review only | not stated | not extracted yet | not extracted yet |
| Zheng, 2017 matched by author and year (probable) | this review only | not stated | not extracted yet | not extracted yet |