To assess whether salivary bacteria, particularly Porphyromonas gingivalis, are associated with clinically confirmed intra-oral halitosis independently of periodontal status.
Approach:
Study design and participants: Cross-sectional study of 102 adults recruited at one dental hospital in Seoul, South Korea; 30 had concordant self-reported and clinician-confirmed halitosis, and 72 had neither.
Microbial analysis: Researchers analyzed unstimulated saliva using 16S rRNA gene sequencing and qPCR, then compared single-species and multi-species classification models using repeated internal cross-validation.
Additional analyses: Analyses examined periodontal status, volatile sulfur compounds, and demographic and behavioral characteristics.
Key Findings:
P gingivalis was detected in all 30 participants with halitosis and 27 of 72 without it; qPCR also found higher median P gingivalis loads in the halitosis group (77,800 vs 10,000 copies/mL).
A P gingivalis-only model had a cross-validated AUROC of 0.913, balanced accuracy of 83.3%, sensitivity of 83.1%, and specificity of 83.5%.
Adding other bacterial species produced numerically higher AUROCs, but improvements were not statistically significant after multiple-comparison adjustment; models without P gingivalis performed near chance.
The association persisted after adjustment for periodontitis and number of teeth. Adding periodontal characteristics changed AUROC by less than 0.01.
P gingivalis abundance correlated with volatile sulfur compounds across the full study population, but these associations were not observed within the halitosis group alone.
Interpretation:
The authors reported that salivary P gingivalis was associated with clinically confirmed intra-oral halitosis independently of periodontal status. They described it as a potential marker whose clinical utility remains unestablished.
Limitations:
Single-center, cross-sectional design and modest sample size limit generalizability and do not establish causality.
Models had internal cross-validation but no external validation; co-occurrence groups were identified in the same cohort used for classification.
Excluding participants with discordant self-reported and clinician assessments may limit applicability to broader populations.
Sequencing-based species identification was provisional, microbial functional activity was not assessed, and the periodontally healthy subgroup included only five participants with halitosis.
Correlation analyses cannot establish biological interactions between co-occurring species.
Conclusion:
The study found that models combining P gingivalis with other taxa added little discriminatory information beyond P gingivalis alone.
A small observational study in collegiate football players found microbiome associations after nonconcussive head impacts, though findings were limited by severe underpowering and high attrition