Cognitive resilience in preclinical Alzheimer's disease: Higher NPTX2 and VGF levels are associated with reduced cognitive decline.
Level 3 - non-randomized controlled study
Longitudinal prospective cohort study with long-term follow-up
PubMed 40945027 · doi:10.1016/j.neurobiolaging.2025.09.002
What was done
The study evaluated 269 initially cognitively unimpaired participants from the BIOCARD cohort (mean baseline age 57.7 years, 59% female) with a mean follow-up of 16 years. Baseline cerebrospinal fluid (CSF) levels of three synaptic proteins—neuronal pentraxin 2 (NPTX2), neurosecretory protein VGF, and AMPA glutamate receptor subunit 4 (GluA4)—were measured using parallel reaction monitoring mass spectrometry. Baseline AD pathology biomarkers (Aβ 1-42/Aβ 1-40 and p-tau 181) were measured using Lumipulse G1200 assays. Linear mixed-effects models tested associations between baseline synaptic proteins and rates of change in global, executive function, and episodic memory composite scores, adjusting for AD biomarker levels.
What was found
Lower baseline Aβ 1-42/Aβ 1-40 and higher p-tau 181 were associated with greater cognitive decline. Controlling for AD pathology, higher baseline NPTX2 and VGF levels were associated with less decline in global cognition (both p <= 0.002) and episodic memory (both p < 0.03). Higher baseline VGF was associated with less executive function decline (p = 0.03), with a similar trend for NPTX2 (p = 0.06). GluA4 showed no significant association with cognitive level or decline across any composite score (all p > 0.08). The abstract did not report effect sizes, beta coefficients, or confidence intervals.
Why it matters
This study shows that baseline CSF levels of NPTX2 and VGF correlate with long-term resistance to cognitive decline despite underlying Alzheimer's pathology. This highlights synaptic integrity markers as potential prognostic indicators and targets for interventions aimed at preserving cognitive function in preclinical AD.
Limits
The abstract reports p-values without effect sizes, beta coefficients, or confidence intervals, preventing assessment of clinical magnitude. The observational design cannot establish causality between higher synaptic protein levels and cognitive resilience. Biomarkers were assessed only at baseline rather than longitudinally, and generalizability beyond the BIOCARD cohort is unestablished.