Within-person variability in kidney measures.

Area of Science:

• Nephrology
• Clinical Chemistry
• Biomarker Analysis

Background:

• Quantifying short-term variability in kidney function measures is crucial for accurate clinical assessment.
• National data from NHANES III provides a robust dataset for evaluating biomarker variability.

Purpose of the Study:

• To quantify short-term within-person variability in standard and non-traditional kidney measures.
• To assess the impact of this variability on the prevalence estimates of reduced kidney function and albuminuria.

Main Methods:

• Analysis of repeated serum and urine samples from NHANES III participants (≥18 years) collected ~18 days apart.
• Calculation of within-person coefficient of variation (CV(w)) for serum creatinine, cystatin C, BTP, B2M, and urine albumin and creatinine.
• Evaluation of variability's effect on estimated glomerular filtration rate (eGFR) and albuminuria prevalence.

Main Results:

• Serum cystatin C (CV(w)=6.8%), creatinine (CV(w)=7.6%), and B2M (CV(w)=8.4%) exhibited low short-term variability.
• Serum BTP showed higher variability (CV(w)=11.6%).
• Urine albumin and creatinine were highly variable (CV(w)>30%), but the albumin-creatinine ratio had lower variability (CV(w)=11.3%).
• Single measurements underestimated eGFR prevalence by ~20% and albuminuria prevalence by 33% compared to repeated testing.

Conclusions:

• Serum creatinine, cystatin C, and B2M are reliable for short-term kidney function assessment due to low variability.
• β-trace protein is more variable than other serum markers.
• Repeated urine assessments are recommended for accurate albuminuria detection due to high variability of single measures.