Work

Association Between Time-Varying Serum Albumin and All-Cause Mortality in Peritoneal Dialysis Patients: A Retrospective Cohort Study Time-Varying Serum Albumin and Mortality in Peritoneal Dialysis

Year: 2026

Type: preprint

Abstract: Background Serum albumin is an established prognostic marker in patients undergoing peritoneal dialysis (PD). However, single baseline measurements may not adequately capture its longitudinal variability. We investigated the association between time-varying serum albumin and all-cause mortality in patients undergoing PD. Methods We conducted a retrospective cohort study of adult patients who initiated PD at our institution. Serial serum albumin measurements were modeled as a time-varying covariate using the Andersen–Gill counting-process formulation of the Cox proportional hazards model. Models were adjusted for age, sex, and diabetes mellitus, with robust sandwich variance estimates clustered by patient. Restricted cubic spline analysis assessed potential nonlinearity. Results Among 1,801 patients with 7,102 laboratory records, 1,680 patients contributing 6,860 observations and experiencing 298 deaths were included in the final analysis. Higher time-varying serum albumin was independently associated with lower all-cause mortality (HR per 1 g/L increase, 0.864; 95% CI, 0.847–0.880; P < 0.001). Older age (HR, 1.040; 95% CI, 1.029–1.050; P < 0.001) and diabetes mellitus (HR, 1.777; 95% CI, 1.395–2.262; P < 0.001) were associated with higher mortality. Restricted cubic spline analysis showed a significant inverse association without evidence of nonlinearity ( P -overall < 0.0001; P -nonlinear = 0.8962). The model had a Harrell C-index of 0.8176. Conclusions Time-varying serum albumin was independently associated with all-cause mortality in patients undergoing PD. Longitudinal albumin assessment may provide prognostic information beyond a single baseline measurement. Trial registration Not applicable.

Source: Research Square

DOI: 10.21203/rs.3.rs-11392988/v1