Biological Variation of Plasma and Urinary Markers of Acute Kidney Injury in Patients with Chronic Kidney Disease

Carter, Joanne L. and Parker, C.T. and Stevens, Paul E. and Eaglestone, Gillian and Knight, Sarah and Farmer, Christopher K. and Lamb, Edmund J. (2016) Biological Variation of Plasma and Urinary Markers of Acute Kidney Injury in Patients with Chronic Kidney Disease. Clinical Chemistry, . ISSN 0009-9147. E-ISSN 1530-8561. (doi: (Full text available)

PDF - Author's Accepted Manuscript
Download (273kB) Preview
Official URL


BACKGROUND: Identification of acute kidney injury (AKI) is predominantly based on changes in plasma creatinine concentration, an insensitive marker. Alternative biomarkers have been proposed. The reference change value (RCV), the point at which biomarker change can be inferred to have occurred with statistical certainty, provides an objective assessment of change in serial tests results in an individual. METHODS: In 80 patients with chronic kidney disease, weekly measurements of blood and urinary biomarker concentrations were undertaken over 6 weeks. Variability was determined and compared before and after adjustment for urinary creatinine and across subgroups stratified by level of kidney function, proteinuria, and presence or absence of diabetes. RESULTS: RCVs were determined for whole blood, plasma, and urinary neutrophil gelatinase-associated lipocalin (111%, 59%, and 693%, respectively), plasma cystatin C (14%), creatinine (17%), and urinary kidney injury molecule 1 (497%), tissue inhibitor of metalloproteinases 2 (454%), N-acetyl-?-d-glucosaminidase (361%), interleukin-18 (819%), albumin (430%), and ?1-microglobulin (216%). Blood biomarkers exhibited lower variability than urinary biomarkers. Generally, adjusting urinary biomarker concentrations for creatinine reduced (P < 0.05) within-individual biological variability (CVI). For some markers, variation differed (P < 0.05) between subgroups. CONCLUSIONS: These data can form a basis for application of these tests in clinical practice and research studies and are applicable across different levels of kidney function and proteinuria and in the presence or absence of diabetes. Most of the studied biomarkers have relatively high CVI (noise) but also have reported large concentration changes in response to renal insult (signal); thus progressive change should be detectable (high signal-to-noise ratio) when baseline data are available.

Item Type: Article
Subjects: R Medicine > RZ Other systems of medicine
Divisions: Faculties > Social Sciences > School of Social Policy Sociology and Social Research > Centre for Health Services Studies
Depositing User: Tony Rees
Date Deposited: 06 May 2016 09:24 UTC
Last Modified: 28 Mar 2017 23:00 UTC
Resource URI: (The current URI for this page, for reference purposes)
  • Depositors only (login required):


Downloads per month over past year