486 - Serum Albumin and BMI Modify Dyslipidemia-Related Mortality in Pediatric Patients Receiving Dialysis

Publication Number: 2473.486

Edward Zitnik, University of Connecticut School of Medicine, West Hartford, CT, United States; Jake Vercellino, Indiana University School of Medicine, San Francisco, CA, United States; Marciana Laster, Riley Hospital for Children at Indiana University Health, Indianapolis, IN, United States

Background: Cardiovascular disease (CVD) is the leading cause of death for patients with end stage kidney disease (ESKD). CVD risk is likely from vascular calcification secondary to mineral bone disorder with a possible contribution from atherosclerosis secondary to dyslipidemia. Data conflict across studies with some showing that dyslipidemia increases mortality risk while others suggest a protective effect, possibly as a surrogate marker for nutritional status. The prevalence and pattern of dyslipidemia in pediatric patients treated with dialysis have been defined; however, no studies have examined the impact of lipid abnormalities on mortality. This study also probes how this association varies based on nutritional status using low serum albumin and underweight BMI as surrogate markers for poor nutritional status.

Objective: To investigate the relationship between dyslipidemia and mortality in children receiving maintenance dialysis, and to explore whether markers of malnutrition modify this association.

Design/Methods: Cox proportional hazard regression models were performed to determine the influence of lipid levels on all-cause mortality for 2,355 pediatric patients ≤21 years old on maintenance dialysis. The fully adjusted model controlled for age, sex, race, etiology of ESKD, dialysis modality, BMI category, and serum albumin. Survival analysis was performed for each lipid parameter and presented as 1st and 3rd quartile lipid level modeled for probability of survival over time. Survival curves were also generated to display BMI categories and low vs. normal serum albumin modeled for probability of survival over time.

Results: Higher Total Cholesterol (TC), LDL-C, and HDL-C had higher probability of survival over time. Triglyceride (TG) showed no change in probability of survival. Underweight BMI had the lowest probability of survival than other BMI groups. Healthy BMI had the highest probability of survival. Low serum albumin had a lower probability of survival than normal serum albumin. Older age, higher albumin, Hispanic race, and peritoneal dialysis were associated with lower risk of mortality in the hazard regression models.

Conclusion(s): Higher lipid levels (TC, LDL-C, HDL-C) are associated with higher probability of survival over time in pediatric patients treated with dialysis. This paradoxical result may suggest that lipid levels are a surrogate marker of nutritional status, as underweight BMI and low serum albumin were associated with a lower probability of survival.

Cohort characteristics
Categorical values are presented as frequency (percentages) and compared using Chi-square test. Continuous variables are presented as the median (interquartile range).

Model-Adjusted Survival Curve by Lipid Level
Survival curves controlled for clinical and demographic characteristics held at typical values (age at median; sex, race, ESKD etiology, dialysis modality at modal levels; BMI at modal group; albumin at median). Red curve represents 1st quartile for the lipid level and blue curve represents the 3rd quartile of the lipid level and modeled for survival probability over time. a) TC, b) LDL-C, c) HDL-C, d) TG.

Model-Adjusted Survival Curve by BMI Category and Serum Albumin
Survival curves controlled for clinical and demographic characteristics held at typical values (age at median; sex, race, ESKD etiology, dialysis modality at modal levels; BMI at modal group for albumin curve; albumin at median for BMI curve). a) red curve represents underweight, green curve represents healthy weight, blue curve represents overweight, and purple curve represents obesity modeled for survival probability over time. b) red curve represents normal serum albumin ≥3.5 g/dL and blue curve represents low serum albumin <3.5 g/dL modeled for survival probability over time.