Background: Anthropometric measurements are simple and reachable tools for self-evaluating and screening patients with a high risk of obstructive sleep apnea (OSA). However, the accumulated relationship of obesity on the anthropometric characteristics of OSA is not well understood. The aim of the study was to show the time-dependent trend of OSA patients and compare overall and regional anthropometric between two ethnicities. Methods: A cumulative meta-analysis was performed to assess obesity metrics in patients with and without OSA between Asians and Caucasians. We searched PubMed, Web of Science, Embase, and Scopus up to Jun 2020. Included studies used body mass index (BMI), neck circumference (NC), waist circumference (WC) and waist-to-hip ratio (WHR) as measures of anthropometric features in the adult OSA population and controls, utilized in-lab polysomnography or home sleep testing with apnea-hypopnea index (AHI) or respiratory disturbance index (RDI) classification, reported ethnicity/race, and were published in English. Any studies lacking one of these criteria or sufficient data were excluded. Results: Forty studies with a total of 19,142 subjects were investigated. Comparison of changes between patients with and without OSA showed that OSA patients had a higher BMI [mean difference (MD) 3.12, 95% confidence interval (CI): 2.51-3.73], NC (MD 3.10, 95% CI: 2.70-3.51), WC (MD 9.84, 95% CI: 8.42-11.26) and waist-hip ratio (MD 0.04, 95% CI: 0.03-0.05) than the control subjects. The accumulated time-dependent increase in population with OSA was significantly apparent with all anthropometric features. BMI increased from 2000 (MD 0.50) to 2012 (MD 3.08-3.48) and remained stable afterwards (MD 2.70-3.17), NC increased from 2000 (MD 0.40) to 2013 (MD 3.09) and remained stable afterwards too (MD 3.06-3.21). WC increased from 2000 (MD 2.00) to 2012 (MD 9.37-10.03) and also remained stable afterwards (MD 8.99-9.84). WHR was stable from 2000 to 2004 with an MD of 0.01 and then stable from 2007 onwards with an increased MD of 0.03-0.04. Compared with Caucasian patients, Asian patients had lower obesity relevant variates. Conclusions: BMI, NC, WC and WHR are associated with OSA in both ethnic groups. Anthropometry for overall and regional obesity could facilitate differentiation of patients with OSA from individuals without OSA by ethnicity.
