Clustering of chronic rhinosinusitis symptomatology reveals novel associations with objective clinical and demographic characteristics.

BACKGROUND: Chronic rhinosinusitis (CRS) is associated with varied head and neck symptomatology and quality-of-life impairments that are not necessarily correlated with each other or with objective measures of disease.
OBJECTIVE: To determine how clustering patterns of CRS symptoms associate with objective clinical findings.
METHODS: Symptom scores from 193 Sinonasal Outcomes Test-22 (SNOT-22) questionnaires, from 177 consecutive CRS patients, were analyzed by principal component analysis (PCA) to uncover fewer and physiologically understandable latent components. Univariate and multivariate regressions were made with patients' demographic characteristics, nasal polyposis, comorbid allergic rhinitis, asthma, gastro-esophageal reflux disease (GERD) or depression, and Lund-Mackay scoring of sinus computed tomography (CT) results.
RESULTS: Four principal components (PCs), heavily weighted on sleep symptoms, nasal symptoms, otologic symptoms, and emotional function symptoms, respectively, are found to primarily describe the variability in patients' SNOT-22 scores. SNOT-22 subset scores reflecting sleep, nasal, otologic, and emotional function symptoms were constructed from corresponding PCs. Only female gender associated with the total SNOT-22 score (p = 0.004), whereas only Lund-Mackay score associated with the nasal subset score (p = 0.015). Allergic rhinitis only associated with the otologic subset score (p = 0.005), whereas only asthma associated with the emotional function subset score (p = 0.027). None of the measured covariates were associated with the sleep subset score.
CONCLUSION: Variability in SNOT-22 scores from CRS patients may be explained by the independent presence of sleep, nasal, otologic, and emotional function symptoms, with which we find novel clinical and demographic associations. These findings may represent clinical evidence for distinct pathophysiologic processes that differentially cause specific CRS symptomatology.