Objective: Aim was to analyze the diagnostic ability of cumulative risk score (CRS), which uses salivary levels of Porphyromonas gingivalis, interleukin (IL)-1β, and matrix metalloproteinase (MMP)-8 in an adaptive design, compared to previously reported thresholds of each marker alone.

Materials and Methods: Oral and general health information of 463 participants were included in the analysis. Having the percentage of bleeding on probing (BOP) > 25%, having at least two sites with probing pocket depth (PPD) of 4–5 mm or having at least one tooth with alveolar bone loss (ABL) of at least 1/3 of the root length were accepted as outcome variables. Being above the salivary threshold concentrations of P. gingivalis, IL-1β, and MMP-8 and CRS values were used as explanatory variables. Receiver operating characteristics (ROC) producing an area under the curve (AUC) and multinomial regression analysis were used in statistical analysis.

Results: CRS provided AUCs larger than any other tested biomarker threshold. Sensitivity and specificity of CRS for detecting clinical markers of periodontitis were acceptable, and a strong association was observed between the highest CRS score and having at least two sites with PPD of 4–5 mm.

Conclusion: CRS brings additional power over fixed thresholds of single biomarkers in detecting periodontitis.