AUTHOR=Polymeris Alexandros A. , Helfenstein Fabrice , Benkert Pascal , Aeschbacher Stefanie , Leppert David , Coslovsky Michael , Willemse Eline , Schaedelin Sabine , Blum Manuel R. , Rodondi Nicolas , Reichlin Tobias , Moschovitis Giorgio , Wuerfel Jens , De Marchis Gian Marco , Engelter Stefan T. , Lyrer Philippe A. , Conen David , Kühne Michael , Osswald Stefan , Bonati Leo H. , Kuhle Jens ,  the Swiss-AF Investigators 

TITLE=Renal Function and Body Mass Index Contribute to Serum Neurofilament Light Chain Levels in Elderly Patients With Atrial Fibrillation

JOURNAL=Frontiers in Neuroscience

VOLUME=16

YEAR=2022

URL=https://www.frontiersin.org/journals/neuroscience/articles/10.3389/fnins.2022.819010

DOI=10.3389/fnins.2022.819010

ISSN=1662-453X

ABSTRACT=<sec><title>Objective</title><p>Serum neurofilament light chain (sNfL) is increasingly used as a neuroaxonal injury biomarker in the elderly. Besides age, little is known about how other physiological factors like renal function and body mass index (BMI) alter its levels. Here, we investigated the association of estimated glomerular filtration rate (eGFR) and BMI with sNfL in a large sample of elderly patients with atrial fibrillation (AF).</p></sec><sec><title>Methods</title><p>This is a cross-sectional analysis from the Swiss-AF Cohort (NCT02105844). We measured sNfL using an ultrasensitive single-molecule array assay. We calculated eGFR using the chronic kidney disease epidemiology collaboration (CKD-EPI) creatinine (eGFR<sub>crea</sub>) and creatinine–cystatin C (eGFR<sub>crea–cys</sub>) formulas, and BMI from weight and height measurements. We evaluated the role of eGFR and BMI as determinants of sNfL levels using multivariable linear regression and the adjusted R<sup>2</sup> (R<sup>2</sup><sub>adj</sub>).</p></sec><sec><title>Results</title><p>Among 2,277 Swiss-AF participants (mean age 73.3 years), eGFR<sub>crea</sub> showed an inverse curvilinear association with sNfL after adjustment for age and cardiovascular comorbidities. BMI also showed an independent, inverse linear association with sNfL. The R<sup>2</sup><sub>adj</sub> of models with age, eGFR<sub>crea</sub>, and BMI alone was 0.26, 0.35, and 0.02, respectively. A model with age and eGFR<sub>crea</sub> combined explained 45% of the sNfL variance. Sensitivity analyses (i) further adjusting for vascular brain lesions (<italic>N</italic> = 1,402 participants with MRI) and (ii) using eGFR<sub>crea–cys</sub> yielded consistent results.</p></sec><sec><title>Interpretation</title><p>In an elderly AF cohort, both renal function and BMI were associated with sNfL, but only renal function explained a substantial proportion of the sNfL variance. This should be taken into account when using sNfL in elderly patients or patients with cardiovascular disease.</p></sec>