Greater arterial stiffness was linked to stronger associations between plasma neurofilament light and cognition among cognitively unimpaired older patients.
In baseline data from 570 patients aged 65 to 80 years, higher neurofilament light (NfL) was associated with lower scores across episodic memory, working memory, processing speed, visuospatial ability, and executive function. Higher carotid–femoral pulse wave velocity (cfPWV) was associated with lower episodic memory, working memory, and processing speed. NfL and cfPWV were not correlated, but interaction testing showed that cfPWV modified the NfL–cognition association: in the higher-cfPWV group, higher NfL related to lower episodic and working memory scores. Patients "with higher and lower cfPWV had comparable NfL levels suggesting similar degrees of axonal injury," wrote the researchers, led by Amani M. Norling, PhD, of The Hinda and Arthur Marcus Institute for Aging Research and the Department of Medicine at Harvard Medical School, both in Boston. "However," they continued, "cognitive performance was worse among those with high cfPWV and high NfL, signifying that arterial stiffness may exacerbate the cognitive impact of axonal damage by further compromising the integrity of brain networks."
They also noted that neither corresponding interactions for processing speed, visuospatial ability, and executive/attentional control, nor interactions using pulse pressure or mean arterial pressure were statistically significant. "Linear modeling may obscure threshold-dependent effects, particularly at higher stiffness levels," they wrote, and arterial stiffness and NfL levels may already be elevated in patients with higher cfPWV. Variability and a potential ceiling effect could also be challenges for further associations. Indeed, they continued, they found a significant negative correlation among patients with a lower cfPWV, which suggested a threshold or nonlinear effect, and that small increases in cfPWV could reflect early arterial changes in concurrent neurodegenerative processes captured by NfL in patients with relatively healthy vascular profiles.
Physical activity may also buffer neuroaxonal injury even in the absence of NfL reduction, Dr. Norling and colleagues wrote, in reference to reports of up to 36% slower cognitive decline in older adults with elevated NfL who engage in regular physical activity.
No three-way interactions with sex or APOE ε4 were detected. Sensitivity analyses were consistent: modeling cfPWV continuously yielded a nonsignificant interaction for working memory; Johnson–Neyman analysis indicated significance for working memory when log NfL exceeded 3.32; and cfPWV stratified at the 50th, 75th, and 90th percentiles showed progressively stronger NfL–working memory slopes. "For working memory," the authors wrote, "cognitive performance peaked near 6.39 m/s with an inflection at 9.55 m/s. Episodic memory showed an inflection at 10.57 m/s."
Investigators analyzed baseline data from a multicenter randomized clinical trial of cognitively unimpaired, community-dwelling older adults across US sites. Plasma NfL was measured on a high-sensitivity single-molecule immunoassay (SIMOA HD-X), and arterial stiffness was assessed by applanation tonometry (SphygmoCor XCEL) to derive cfPWV; patients were grouped by age-adjusted cut points (~9.7 m/s for ages 60 to 69; 10.6 m/s for patients older than 70). Cognitive tests across 5 domains were combined into composite scores, and models were adjusted for age, sex, education, site, and waist circumference. The analytic sample included 144 higher-stiffness and 426 lower-stiffness patients; NfL levels were similar between groups. Participants who were excluded for missing data had higher BMI and lower cognitive performance than those included.
The analysis had several limitations. The cross-sectional design did not determine temporal ordering among arterial stiffness, NfL, and cognitive performance. The cohort largely comprised relatively healthy, community-dwelling older adults, which may limit generalizability. Cerebral blood flow and other vascular measures were not collected, restricting mechanistic interpretation. Exclusions related to missing data may have introduced bias.
Prospective follow-up from this cohort is needed to evaluate temporal relationships and whether trajectories of cfPWV and NfL track subsequent cognitive change. Clinical implications were not established in this baseline analysis.
The authors reported no conflicts of interest.
Source: Alzheimer’s Association
