RESUMO
OBJECTIVES: This retrospective cohort study investigated the association of diabetes with mortality in hemodialysis patients with regard to obesity, sarcopenia, and sarcopenic obesity, along with examining the prevalence of each group and diabetes. METHODS: Muscle strength, muscle mass, and fat mass were evaluated using a hand dynamometer and dual-energy X-ray absorptiometry, respectively, in 308 chronic hemodialysis patients (age 58.0 ± 11.9 years, hemodialysis duration 6.5 ± 6.0 years, males 60.1%, diabetes 32.8%). Sarcopenia was defined according to the new criteria established by the Asian Working Group on Sarcopenia 2019. Obesity was defined by percent body fat mass (males ≥25%, females ≥35%). RESULTS: The enrolled patients were divided into the normal (38.7%), obesity (18.8%), sarcopenia (26.9%), and sarcopenic obesity (15.6%) groups. The prevalence of diabetes was significantly skewed among the 4 groups (χ2 test, P = .0057), being higher in the sarcopenic obesity group (54.2%) compared to the others (25.9-33.7%). Multivariate regression analysis revealed that diabetes was significantly and independently associated with sarcopenic obesity (odds ratio 3.495, 95% confidence interval 1.683-7.255, P = .0008) after adjustments for several cofounders, but not significantly associated with sarcopenia. During the follow-up period of 76 ± 35 months, 100 patients died. Those in the sarcopenia and sarcopenic obesity groups had significantly higher rates of all-cause mortality compared to patients in the normal and obesity groups (P = .0004, log-rank test). Furthermore, multivariate Cox proportional hazards analysis revealed that presence of diabetes was significantly associated with higher all-cause mortality in all 308 patients, after adjustments for several factors, including the presence of each group in 4 models. CONCLUSION: Sarcopenic obesity is highly prevalent in chronic hemodialysis patients. Diabetes was found to be a significant and independent contributor to the presence of sarcopenic obesity. Diabetes was shown to be a significant predictor of all-cause mortality, independent of the present normal, obesity, sarcopenia, and sarcopenic obesity groups.