Renal function affects circulating levels of neurofilament light chain: MarkVCID‐CHARGE

Abstract

AbstractBackgroundPlasma NfL is a promising fluid biomarker for risk stratification and quantifying the extent of neuronal injury in conditions such as vascular cognitive impairment and dementia (VCID) and Alzheimer’s dementia. However, systemic factors such as kidney function can affect circulating levels and this requires further investigation to ensure appropriate use as a biomarker.MethodWe examined the cross‐sectional association between plasma NfL levels and renal function among adult participants from four community‐based studies in the MarkVCID‐Cohorts for Heart and Aging Research in Genomic Epidemiology consortium who were free from stroke, multiple sclerosis, dementia, and renal failure. Estimated glomerular filtration rate (eGFR, mL/min/1.73m2) was calculated using the CKD‐EPI equation and categorized using Kidney Disease: Improving Global Outcomes (KDIGO) classification. Plasma NfL was measured using single molecule array technology. Multivariable linear regression models were used to investigate the linear relationship between eGFR and NfL, overall and stratified by KDIGO category. Study level estimates were combined in meta‐analysis. Models were adjusted for age, sex, sub‐cohort, hypertension, diabetes, body mass index, smoking, and race. Factors affecting eGFR can also cause MRI brain injury, hence we examined whether any apparent increase in NfL levels in persons with a lower eGFR merely reflected greater brain injury by performing a sensitivity analysis in a subsample with brain MRI, additionally adjusting in a single model for brain volume, white matter hyperintensities, and free water.ResultIn meta‐analysis (n = 5,954), lower eGFR was associated with higher plasma NfL (β = ‐0.0073, standard error[SE] = 0.0021, P = 0.0006). This relationship persisted within categories of reduced renal function, eGFR 60‐74 (mild), 45‐59 (moderate), and 15‐44 (severe) (P‐values<0.0001), but was not evident with normal eGFR ≥90 (P = 0.3) or mildly decreased eGFR 75‐89 (P = 0.2). Further adjustment for white matter hyperintensities, brain volume, and free water among a subgroup with brain MRI (n = 2192) did not change the associations of eGFR and NfL (β = ‐0.0065,SE = 0.0025, P = 0.0091).ConclusionReduced renal function is associated with higher circulating NfL, even at an eGFR range (60‐74) considered normal in clinical practice. Adjusting for renal function may improve the accuracy of plasma NfL as a biomarker for risk stratification or for assessment of neuronal injury in conditions such as VCID.