Crude and multivariate Cox regression analyses.

Obesity and adipose tissue are commonly regarded as detrimental factors linked to adverse outcomes, including cardiovascular and metabolic diseases. However, the obesity paradox is obesity that may provide survival benefits for chronic diseases including patients undergoing hemodialysis. Fat mass can be a surrogate marker for nutrition status in patients undergoing hemodialysis. Thus, this study evaluated subcutaneous fat and all-cause mortality in patients initiating hemodialysis. A total of 123 patients initiating hemodialysis were included in this study. MATLAB (version R2014a) was used to identify subcutaneous fat area (SFA) and visceral fat area (VFA) in computed tomography images for the analysis of body composition. The survival rate was calculated using Cox regression analysis. The Kaplan–Meier survival rates were 70.0% and 85.7% in the low and high subcutaneous fat area (SFA) groups, respectively (log rank, p = 0.021). In Cox analysis, the low SFA group showed high risk for all-cause mortality than the high SFA group (hazard ratio (HR) 3.541, 95% CI 1.358–9.235, p = 0.010). In subgroup univariate analysis, the risk for all-cause mortality was higher in patients with low SFA and diabetes than those with high SFA and diabetes (HR 3.541, 95% CI 1.358–9.235, p = 0.010). In multivariate analysis, the risk for all-cause mortality was higher in patients with low SFA and diabetes than those with high SFA and diabetes (HR 4.615, 95% CI 1.484–14.351, p = 0.008). Conclusively, low SFA increases the risk of 2-year all-cause mortality, and SFA analysis can provide information for risk evaluation for patients initiating hemodialysis.