Defining Extreme Phenotypes of OSA Across International Sleep Centers
Fabiola G. Rizzatti,Diego R. Mazzotti,Jesse Mindel,Greg Maislin,Brendan T. Keenan,Lia Bittencourt,Ning-Hung Chen,Peter A. Cistulli,Nigel McArdle,Frances M. Pack,Bhajan Singh,Kate Sutherland,Bryndis Benediktsdottir,Ingo Fietze,Thorarinn Gislason,Diane C. Lim,Thomas Penzel,Bernd Sanner,Fang Han,Qing Yun Li,Richard Schwab,Sergio Tufik,Allan Pack,Ulysses J. Magalang
DOI: https://doi.org/10.1016/j.chest.2020.03.055
IF: 9.6
2020-01-01
Chest
Abstract:BACKGROUND: Extreme phenotypes of OSA have not been systematically defined. RESEARCH QUESTION: This study developed objective definitions of extreme phenotypes of OSA by using a multivariate approach. The utility of these definitions for identifying characteristics that confer predisposition toward or protection against OSA is shown in a new prospective sample. STUDY DESIGN AND METHODS: In a large international sample, race-specific liability scores were calculated from a weighted logistic regression that included age, sex, and BMI. Extreme cases were defined as individuals with an apnea-hypopnea index (AHI) >= 30 events/hour but low likelihood of OSA based on age, sex, and BMI (liability scores 90th percentile). Similarly, extreme controls were individuals with an AHI < 5 events/hour but high likelihood of OSA (liability scores < 10th percentile). Definitions were applied to a prospective sample from the Sleep Apnea Global Interdisciplinary Consortium, and differences in photography based craniofacial and intraoral phenotypes were evaluated. RESULTS: This study included retrospective data from 81,338 individuals. A total of 4,168 extreme cases and 1,432 extreme controls were identified by using liability scores. Extreme cases were younger (43.1 +/- 14.7 years), overweight (28.6 +/- 6.8 kg/m(2)), and predominantly female (71.1%). Extreme controls were older (53.8 +/- 14.1 years), obese (34.0 +/- 8.1 kg/m(2)), and predominantly male (65.8%). These objective definitions identified 29 extreme cases and 87 extreme controls among 1,424 Sleep Apnea Global Interdisciplinary Consortium participants with photography-based phenotyping. Comparisons suggest that a greater cervicomental angle increases risk for OSA in the absence of clinical risk factors, and smaller facial widths are protective in the presence of clinical risk factors. INTERPRETATION: This objective definition can be applied in sleep centers throughout the world to consistently define OSA extreme phenotypes for future studies on genetic, anatomic, and physiologic pathways to OSA.