Anthropometrical phenotypes are important when explaining obstructive sleep apnea in female bariatric cohorts

• DOI: 10.1111/jsr.12830
• Publication Date: 2019-02-11T08:20:25Z
• Authors: Mercè Gasa; Carla López‐Padrós; Carmen Monasterio; Neus Salord; Mercedes Mayos; Núria Vilarrasa; Fernando Fernande...; Josep M. Montserrat; Jordi Dorca
• Keywords: Adult; 2. Zero hunger; Sleep Apnea, Obstructive; Polysomnography; body mass index; neck circumference; waist circumference; morbid obesity; Obesity, Morbid; 03 medical and health sciences; Apnea-hypopnea index; female; Bariatrics; Cross-Sectional Studies; Phenotype; 0302 clinical medicine; Risk Factors; Humans; Female; Prospective Studies

AbstractCentral obesity is the main risk factor for obstructive sleep apnea (OSA). Whether there exists a central‐obesity anthropometric that better explains apnea–hypopnea index (AHI) variability in the general population and in sleep cohorts is unknown, and this is even less explored among increasing grades of obesity. The objective of the study is to investigate whether there is an anthropometric that better explains AHI variability in a sample of morbidly obese women awaiting bariatric surgery (BS). A prospective multicentre cross‐sectional study was conducted in consecutive women before BS. Demographic and anthropometric characteristics included age, body mass index (BMI), neck circumference (NC), waist circumference (WC), hip circumference (HC) and waist‐to‐hip ratio (WHR). OSA was diagnosed by polysomnography. The capacity of anthropometrics to explain AHI variance was investigated using regression linear models. A total of 115 women were evaluated: age, 44 ± 10 years; BMI, 46 ± 5 kg/m2; AHI, 35 ± 26 events/hr. AHI was associated with all anthropometrics except weight, height and HC. The best univariate predictor was WHR, which accounted for 15% of AHI variance. The simplest model (age + BMI) accounted for 9%, which increased to 20% when applying more complex measurements (age + BMI + NC + WC + HC). The explanatory capacity did not change significantly when applying a simpler model (age + WHR + NC, 19%). In this female morbidly obese cohort, anthropometrics explained one‐fifth of AHI variability. WHR is the best univariate parameter and models including waist and neck data provide more information than BMI when explaining AHI variability. Thus, even in young women with extreme obesity, OSA seems to be linked to a specific central‐obesity phenotype rather than to a whole‐obesity pattern.