LAUSR

OBJECTIVE this retrospective study aims to compare the prevalence and diagnostic agreement of sarcopenic obesity (SO) using different obesity diagnostic methods among Chinese community-dwelling older adults. METHODS SO was diagnosed with sarcopenia and obesity diagnostic methods. Sarcopenia was defined using the Asian Working Group for Sarcopenia criteria 2019 (AWGS2019). Four widely used indicators were used to define obesity: body mass index (BMI), waist circumference (WC), percent of body fat (PBF) and visceral fat area (VFA). Cohen's kappa was used to analyse the diagnosis agreement of SO between different diagnostic methods. RESULTS a total of 1,050 participants were included, including 347 men (71.3 ± 7.4 years) and 703 women (69.9 ± 7.5 years). The prevalence of sarcopenia was 25% in total participants, there was no difference between men (24.2%) and women (25.5%), (P = 0.705). With different obesity diagnostic methods, the obesity prevalence ranged from 4.1 to 42.2%, the SO prevalence was 0.1-7.9%. The diagnosis agreement of SO was poor-to-moderate (κ ranged from -0.002 to 0.682). Among the four diagnostic methods, AWGS combined with BMI had the poorest agreement (κ = -0.002 with other methods), AWGS combined with VFA had the best agreement (κ = 0.641 and 0.682 with AWGS combined with PBF and with AWGS combined with WC, respectively). CONCLUSION the prevalence of SO vary considerably and the diagnostic agreement is poor-to-moderate with non-uniform diagnostic methods. BMI has the lowest sensitivity, whereas VFA has the highest sensitivity in diagnosis of SO, and VFA has a relatively good diagnostic agreement with other diagnostic methods.