[Usefulness of cephalometry and pharyngeal findings in the primary diagnosis of obstructive sleep apnea syndrome]

We studied the usefulness of cephalometry and pharyngeal findings in determining efficient primary diagnosis of obstructive sleep apnea syndrome (OSAS). Subjects were 171 adult men with sleep-disordered breathing, who underwent cephalometry and polysomnography (PSG) and had pharyngeal findings evaluated by modified Mallampati grade (MMP) and palatine tonsil size. An apnea-hypopnea index (AHI) ofor = 15 in PSG was considered to indicate OSAS. We analyzed the correlation between AHI and other parameters and conducted stepwise multiple regression analysis to predict AHI, and studied the screening performance of prediction equations using a receiver operating characteristic (ROC) curve. Of the 8 cephalometric parameters examined, the length of the soft palate (PNS-P; p = 0.011) and the distance from the mandibular plane to the hyoid bone (MP-H; p0.001) correlated significantly with AHI. Two indices of the pharyngeal finding and body mass index (BMI) also significantly correlated with the AHI (MMP; p0.001, tonsil size; p = 0.005, BMI; p0.001). We next set up three prediction equations including (a) two significant cephalometric parameters (PNS-P and MP-H), (b) two pharyngeal indices (MMP and tonsil size), or (c) both. BMI was included in all three. ROC curve analysis showed that the screening performance of prediction equations was (c)(a)(b). Sensitivity and of equation (c) with a cutoff of 15 were 0.95 and specificity 0.25. These results indicate that both cephalometric parameters and pharyngeal indices should be included in conducting an efficient primary diagnosis for OSAS.