Hidden Oral Inflammation and Biomarker Misclassification in Obstructive Sleep Apnea Trials

Abstract

Inflammatory biomarkers are used as surrogate endpoints in obstructive sleep apnea (OSA) trials, assuming the measured signal originates from the airway or systemic response to intermittent hypoxia. Periodontitis – a common, extra-pulmonary inflammatory reservoir – elevates systemic active matrix metalloproteinase-8 (aMMP-8) independently of OSA. This Perspective examines whether unmeasured oral inflammation confounds biomarker attribution in OSA trials.

Epidemiological and interventional evidence indicates that periodontitis is highly prevalent, is more common in individuals with OSA, and substantially elevates circulating aMMP-8. Moreover, non-surgical periodontal therapy reduces this signal to a degree comparable to early CPAP effects. In published CPAP trials that reported inflammatory or cardiovascular outcomes, periodontal status has not been reported or incorporated into primary analyses. We propose a heuristic conceptual model: measured circulating aMMP-8 = OSA-related component + periodontal-derived component + measurement noise. Without adjustment for periodontal status, observed changes cannot be uniquely attributed to OSA treatment. The ORBIT heuristic framework (Oral Reservoir Bias Inflammatory Trial) is introduced as a conceptual prompt for reporting oral inflammatory burden.

Inflammatory biomarker responses in OSA trials are consistent with a composite signal that includes an unmeasured oral component. A testable prediction is that adjusting for periodontal status would reduce between-trial heterogeneity of MMP-8 effect estimates without altering within-subject respiratory outcomes. Prospective validation is warranted. The framework extends beyond aMMP-8 to any circulating inflammatory marker influenced by extra-target inflammatory reservoirs.