Optimized NMR-based metabolomic protocol for salivary analysis in clinical practice

Saliva remains underexplored in metabolomics compared to widely used biofluids such as blood and urine. However, its non-invasive, rapid, and cost-effective collection, as well as its suitability for self-sampling, make it particularly attractive for clinical and personalized medicine. Moreover, saliva is expected to provide complementary metabolic information to other biofluids. In this context, the present study aims to identify the most suitable sample collection and preparation protocol for maximizing relevant and informative metabolomic data from salivary NMR profiles. Four preparation protocols, selected and adapted from the literature, were systematically compared. Evaluation criteria included the number of identified metabolites, their concentrations, and the intra-day repeatability based on PCA (buckets). The most promising one is an in-house adapted method involving sequential centrifugation, freeze-drying and ultrafiltration (protocol 4). It emerges as the most robust protocol, providing the broadest metabolome coverage with 42 metabolites identified and quantified. This method was further assessed using ANOVA to determine intra- and inter-day precision. Most metabolites demonstrate excellent repeatability (coefficients of variation below 10%) and confirm the reliability of the protocol for quantitative metabolomics analysis. Overall, the optimized approach yields high-quality spectra, wide metabolic coverage and strong analytical precision, making it well suited for reproducible salivary NMR metabolomics. Beyond its methodological contribution, this work highlights saliva as a promising biofluid for diagnostics, disease monitoring and personalized medicine. With appropriate standardization of collection and preparation procedures, saliva could emerge as a biofluid of choice in clinical metabolomics, contributing to advances in preventive and precision healthcare.