Facial and Intraoral Photographic Traits Related to Sleep Apnea in a Clinical Sample with Genetic Ancestry Analysis.
Kate Sutherland,Soriul Kim,Olivia J. Veatch,Brendan T. Keenan,Lia Bittencourt,Ning-Hung Chen,Thorarinn Gislason,Fang Han,Niusha Jafari,Qing Yun Li,Diane C. Lim,Greg Maislin,Ulysses Magalang,Diego R. Mazzotti,Nigel McArdle,Jesse Mindel,Allan I. Pack,Thomas Penzel,Bhajan Singh,Andrew Wiemken,Liyue Xu,Yun Sun,Sergio Tufik,Richard J. Schwab,Peter A. Cistulli
DOI: https://doi.org/10.1513/annalsats.202207-577oc
2023-01-01
Abstract:Rationale: Craniofacial and pharyngeal morphology influences risk for obstructive sleep apnea (OSA). Quantitative photography provides phenotypic information about these anatomical factors and is feasible in large samples. However, whether associations between morphology and OSA severity differ among populations is unknown. Objectives: The aim of this study was to examine this question in a large sample encompassing people from different ancestral backgrounds. Methods: Participants in SAGIC (Sleep Apnea Global Interdisciplinary Consortium) with genotyping data were included (N= 2,393). Associations between photography-based measures and OSA severity were assessed using linear regression, controlling for age, sex, body mass index, and genetic ancestry. Subgroups (on the basis of 1000 Genomes reference populations) were identified: European (EUR), East Asian, American, South Asian, and African (AFR). Interaction tests were used to assess if genetically determined ancestry group modified these relationships. Results: Cluster analysis of genetic ancestry proportions identified four ancestrally defined groups: East Asia (48.3%), EUR (33.6%), admixed (11.7%; 46% EUR, 27% Americas, and 22% AFR), and AFR (6.4%). Multiple anatomical traits were associated with more severe OSA independent of ancestry, including larger cervicomental angle (standardized b [95% confidence interval (CI)] = 0.11 [0.06-0.16]; P < 0.001), mandibular width (standardized b [95% CI] = 0.15 [0.10-0.20]; P < 0.001), and tongue thickness (standardized b [95% CI] = 0.06 [0.02-0.10]; P = 0.001) and smaller airway width (standardized b [95% CI] = 20.08 [20.15 to 20.002]; P = 0.043). Other traits, including maxillary and mandibular depth angles and lower face height, demonstrated different associations with OSA severity on the basis of ancestrally defined subgroups. Conclusions: We confirm that multiple facial and intraoral photographic measurements are associated with OSA severity independent of ancestral background, whereas others differ in their associations among the ancestrally defined subgroups.