Abstract
BACKGROUND: Gait impairment leads to increased mobility decline and may have neurological contributions. This study explores how neurological biomarkers are related to gait in older adults. METHODS: We studied participants in the Cardiovascular Health Study, a population-based cohort of older Americans, who underwent a serum biomarker assessment from samples collected in 1996-1997 for neurofilament light chain (NfL), glial fibrillary acidic protein, ubiquitin carboxy-terminal hydrolase L1, and total tau (n = 1 959, mean age = 78.0 years, 60.8% female). In a subsample (n = 380), cross-sectional associations with quantitative gait measures were explored. This subsample was assessed on a mat for gait speed, step length, double support time, step time, step length variability, and step time variability. Gait speed was also measured over a 15-ft walkway annually from 1996-1997 to 1998-1999 for longitudinal analyses. Linear regression models assessed cross-sectional associations of biomarkers with gait measures, whereas mixed effects models assessed longitudinal gait speed change from baseline to 1998-1999. RESULTS: Neurofilament light chain was significantly associated with annual gait speed decline (standardized β = -0.64 m/s, 95% CI: [-1.23, -0.06]) after adjustment for demographic and health factors. Among gait mat-assessed phenotypes, NfL was also cross-sectionally associated with gait speed (β = 0.001 m/s [0.0003, 0.002]) but not with other gait measures. None of the remaining biomarkers were significantly related to gait in either longitudinal or cross-sectional analyses. CONCLUSIONS: Higher NfL levels were related to greater annual gait speed decline. Gait speed decline may be related to axonal degeneration. The clinical utility of NfL should be explored.