The aim of this study was to investigate the relationship between red blood cell distribution width and albumin ratio (RAR) levels and mortality in adult patients with sarcopenic obesity in the United States. The study included 1,361 adult patients with sarcopenic obesity from the National Health and Nutrition Examination Survey (1999–2006). The X-tile was used to determine the optimal subgroup thresholds for RAR values, and propensity score matching (PSM) was employed to reduce baseline bias. Cox regression analysis, Kaplan-Meier survival curves, and restricted cubic spline analysis were utilized to assess the relationship between RAR levels and all-cause and cardiovascular mortality. Subgroup analysis and the Subpopulation Treatment Effect Pattern Plot were employed to determine survival advantages across different subgroups. Time-dependent ROC analysis to evaluate the accuracy of RAR level in predicting survival outcomes at different time points. Post-PSM multifactorial Cox regression analyses revealed that RAR was a significant independent predictor of all-cause mortality (HR 1.487, 95% CI: 1.259–1.756) and an independent risk factor for cardiovascular mortality (HR 1.487, 95% CI: 1.260–1.758) in patients with sarcopenic obesity. The survival advantage was consistent across subgroups. Restricted cubic spline analysis indicated an approximate S-shaped association between RAR levels and mortality. Time-dependent ROC curves demonstrate that the areas under the all-cause mortality curves at the RAR level for 1-year, 3-year, 5-year, and 10-year are 0.79, 0.66, 0.64, and 0.63, respectively. The areas under the cardiovascular mortality curve are 0.80, 0.70, 0.66, and 0.61, respectively. Moreover, in comparison to the baseline model lacking covariates, the AUC values of the joint model exhibited heightened levels at various time points. Therefore, We demonstrated that the RAR level is an independent prognostic factor for mortality risk in the American population with sarcopenic obesity, and it is reasonable to consider the RAR level as a simple and effective risk prediction tool.
Supplementary Information
The online version contains supplementary material available at 10.1038/s41598-024-79055-1.
