Abstract: TH-PO750
Characterizing Nonlinear GFR Decline in Children with Kidney Diseases
Session Information
- Pediatric CKD
November 07, 2019 | Location: Exhibit Hall, Walter E. Washington Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: Pediatric Nephrology
- 1700 Pediatric Nephrology
Authors
- Ng, Derek, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland, United States
- Warady, Bradley A., Children's Mercy Kansas City, Kansas City, Missouri, United States
- Furth, Susan L., The Children's Hospital of Philadelphia, Philadelphia, Pennsylvania, United States
- Schwartz, George J., University of Rochester, Rochester, New York, United States
Background
GFR decline in kidney disease is frequently treated as linear in clinical research. The Chronic Kidney Disease in Children (CKiD) cohort provided long-term follow-up to characterize and describe GFR decline using non-linear models and accounted for risk stage using the newly developed pediatric classification.
Methods
CKiD participants contributed up to 13 follow-up years and provided annual estimated GFR data. To account for diagnosis, separate models were fit by non-glomerular and glomerular diseases. Linear mixed effects models were fit with log GFR as the outcome and a quadratic effect of time, with random effects for intercept, time and time2. Baseline proteinuria categories (<0.5, 0.5 to 2, >2mg/mgCr) modified each parameter. Empirical Bayes estimates quantified individual-specific entry GFR, and changes within 5 years and between 5 and 10 years and were stratified by baseline CKD risk stage (initial GFR entry and observed proteinuria). Stable GFR was defined as no decline.
Results
A total of 757 and 275 children with non-glomerular and glomerular CKD, contributed 3926 and 1213 observations with 45% and 30% contributing at least 5 years. Most participants entered at Stage A or B (lowest severity) risk: for the non-glomerular and glomerular groups, 48% and 39% were in Stage A; 32% and 31% were in Stage B, respectively. The quadratic term for time offered significantly better fit than a linear effect only, for both groups (both p<0.001), and proteinuria significantly modified GFR trajectory. Nearly all children (95%) with non-glomerular diseases and 74% of those with glomerular diseases experienced faster percent decline in the second 5 years compared to the first 5 years, regardless of initial CKD stage. About 1/3 had stable GFR in the first 5 years in both non-glomerular and glomerular groups (27% and 31%), and Stages A and B comprised the vast majority of these first 5 year stable groups (86% and 69%). However, only 5% and 14% had a stable GFR in the second 5 years, respectively.
Conclusion
In general, in children with kidney diseases, GFR decline was not constant, but accelerated over time, and there was substantially more variability in glomerular disease. Mixed effects models provided individual non-linear trajectories and the challenge of heterogeneous disease duration at study entry was overcome by stratification by baseline CKD risk stage.
Funding
- NIDDK Support