This study aims to summarize recent studies available on untargeted metabolomics employed for periodontitis diagnosis, from saliva and gingival crevicular fluid samples, to identify recurring metabolites with biomarker-value potential. A secondary objective was to analysudurue the protocols of existing studies, to facilitate further research.
Three databases were electronically searched for relevant studies (PubMed, Web of Science, Scopus). Risk of bias assessment was performed using the Newcastle-Ottawa scale (NOS). Data was extracted from studies, regarding general characteristics and conclusions, population characteristics, periodontal protocols, and metabolomics protocols. Metabolic pathway analysis was performed for recurrent metabolites.
After screening 405 studies, 13 studies (10 using saliva samples, 3 using GCF samples) were included. 22 metabolites were identified in more than one study and included into the pathway analysis. Butyrate, lactate, isoleucine, glucose, pyruvate, isovalerate, hypoxanthine/xanthine, proline, valine, phenylalanine, and ethanol were most frequently encountered and were found upregulated in periodontitis patients compared to periodontally healthy patients.
Metabolomics could provide valuable opportunities in validating potential biomarkers or diagnosis panels, contributing to the screening, prognosis, progression and monitoring of periodontitis. Further studies on larger populations and using established protocols are needed. (PROSPERO CRD42023470339).


