Development and internal validation of a preliminary cumulative odontogenic risk categorization score for identifying factors associated with maxillary sinus membrane thickening: a retrospective cross-sectional study

BACKGROUND: Periodontal bone loss (PBL) is a traditional metric for periodontitis, but its ability to reflect the total regional inflammatory burden in the posterior maxilla is limited. This study aimed to develop and internally validate a preliminary cumulative odontogenic score (COS) by integrating periodontal and endodontic parameters to characterize their combined association with maxillary sinus membrane thickening (SMT). METHODS: This retrospective cross-sectional study analyzed 1,056 dentate maxillary sinuses from 562 patients. Odontogenic variables, including PBL severity, endodontic treatment history, and periapical lesions, were assessed using cone-beam computed tomography (CBCT). Multivariable logistic regression using generalized estimating equations (GEE) was performed to identify independent factors associated with SMT (> 2 mm). A COS (range 1-7) was derived from weighted β regression coefficients. Classification performance was evaluated through ROC analysis, calibration plots, and bootstrapping (1,000 resamples). RESULTS: SMT was prevalent in 46.3% of sinuses. PBL, periapical lesions, and endodontic treatment were all independent contributors to SMT (p < 0.001). The COS demonstrated significantly superior classification performance compared to PBL alone (AUC 0.718 vs. 0.602; p < 0.001) and exhibited adequate calibration (slope = 1.00). Internal validation via bootstrapping confirmed the stability of the associations. Each one-point increase in the COS was associated with a 30% increase in the odds of SMT. Using a clinically relevant threshold of COS > 3, the prevalence of SMT in the high-COS group was nearly double that of the low-COS group (59.9% vs. 30.5%). CONCLUSIONS: Integrating periodontal and endodontic conditions into a unified scoring framework provides a more adequate characterization of the dental-sinus association than individual parameters. The internally validated COS offers a structured approach for preoperative radiographic evaluation and can function as a preliminary cross-sectional risk categorization score to characterize interdisciplinary associations in the posterior maxilla.