Can we identify a Surface Phenotype and Prediction Model for Obstructive Sleep Apnoea? A case-Control Study

Obstructive Sleep Apnoea (OSA) remains relatively underdiagnosed and associated with significant comorbidity. The present study aimed to explore the existence of a surface facial-cervical phenotype, prediction model and the presence of a surface marker for adults with OSA. A hospital-based prospective case-control study design was adopted, with 118 middle-aged Caucasian males (56 controls and 62 OSA subjects) recruited. Each subject underwent a clinical examination and overnight sleep study to confirm their grouping. Stereophotography provided a reliable 3-dimensional surface facial-cervical measurement technique, with multilevel statistical analysis performed. A surface facial-cervical & clinical phenotype was identified for OSA, with the predominant characteristics being: a short and enlarged neck circumference (p < 0.001), large mandibular width (p < 0.001), forward head posture (p < 0.001), increased lower facial height [P < 0.002]; increased sleep apnoea clinical scores; high BMI and aggregated Mallampati score (class 3 and 4; p < 0.001).The regression model of surface and clinical variables optimally predicted (area under receiver operator curve, AUC = 0.82), with a high positive likelihood ratio (LR + 6.02). The surface model not only successfully identified OSA subjects from controls (AUC = 0.77) but also presented as a marker. A surface phenotypic pattern, predictive model and marker for OSA in Caucasian men, was identified.