Abstract Background Burning Mouth Syndrome (BMS) is a chronic orofacial pain condition without objective diagnostic criteria and with a substantial impact on quality of life. This umbrella review aimed to evaluate the available evidence on salivary biomarkers in BMS, identify which biomarkers show the most consistent associations with the condition, and assess the methodological quality of existing systematic reviews and meta-analyses. Methods A comprehensive search of PubMed, Embase, Web of Science, Scopus, TRIP, and Google Scholar was conducted without language or date restrictions. Systematic reviews and meta-analyses assessing salivary biomarkers in patients with BMS were included. Of the 561 records screened, six reviews met the eligibility criteria. Methodological quality was assessed using the AMSTAR-2 tool. Results AMSTAR 2 appraisal rated four reviews as moderate confidence and two as low confidence; none was rated high or critically low. The evidence base showed very high overlap among primary studies, indicating that repeated findings across reviews should not be interpreted as fully independent replications. Cortisol showed the most consistent evidence of elevation in BMS. α-amylase and IgA were also frequently elevated, particularly in reviews on stress-related biomarkers. However, their interpretation was limited by methodological heterogeneity, overlap of primary studies and limited disease specificity. Cytokines, sex hormones, opiorphin, trace elements and oxidative stress-related markers showed inconsistent or sparse evidence. Conclusions Salivary cortisol, α-amylase and IgA may reflect stress-related neuroendocrine, autonomic and mucosal immune dysregulation in BMS. However, current evidence remains methodologically limited and does not support the use of any salivary biomarker as a standalone diagnostic test. Clinical relevance Saliva-based biomarkers may support mechanistic phenotyping in future research, but standardised protocols, rigorous study designs and improved patient stratification are required before they can inform clinical decision-making.
Machete et al. (Wed,) studied this question.