Longitudinal plasma phosphorylated-tau217 and other related biomarkers in a non-demented Alzheimer's risk-enhanced sample

Abstract

Introduction: Understanding longitudinal change in key plasma biomarkers will aid in detecting presymptomatic Alzheimer's disease (AD).

Methods: Serial plasma samples from 424 Wisconsin Registry for Alzheimer's Prevention participants were analyzed for phosphorylated-tau217 (p-tau217; ALZpath) and other AD biomarkers, to study longitudinal trajectories in relation to disease, health factors, and cognitive decline. Of the participants, 18.6% with known amyloid status were amyloid positive (A+); 97.2% were cognitively unimpaired (CU).

Results: In the CU, amyloid-negative (A-) subset, plasma p-tau217 levels increased modestly with age but were unaffected by body mass index and kidney function. In the whole sample, average p-tau217 change rates were higher in those who were A+ (e.g., simple slopes(se) for A+ and A- at age 60 were 0.232(0.028) and 0.038(0.013))). High baseline p-tau217 levels predicted faster preclinical cognitive decline.

Discussion: p-tau217 stands out among markers for its strong association with disease and cognitive decline, indicating its potential for early AD detection and monitoring progression.

Highlights: Phosphorylated-tau217 (p-tau217) trajectories were significantly different in people who were known to be amyloid positive. Subtle age-related trajectories were seen for all the plasma markers in amyloid-negative cognitively unimpaired. Kidney function and body mass index were not associated with plasma p-tau217 trajectories. Higher plasma p-tau217 was associated with faster preclinical cognitive decline.

Keywords: ALZpath; BMI; biomarkers; cognitive composite; plasma; preclinical Alzheimer's disease.

FIGURE 1

The spaghetti plot of plasma biomarkers in all available plasma samples aligned with the predicted values with A status. Amyloid positive (A+): Positron emission tomography (PET) Pittsburgh Compound B (PiB) Distribution volume ratio (DVR) >1.19 corresponding to a centiloid of 21.6, or cerebrospinal fluid (CSF) p‐tau/Aβ42 ratio >0.038 when amyloid PET was missing (N = 71); A‐: N = 311; Unknown = 42.

FIGURE 2

Interaction plot from mixed‐effects model of plasma p‐tau217 in cognitively unimpaired amyloid‐negative (top row) and whole sample with known A status (bottom row). Estimates come from the significant main effects or interaction effects of best‐fitting regression models for plasma p‐tau217 in healthy control (A, B, and C), in whole sample with known A status before and after adding A status to the model (D and E), and in sample with PET amyloid (F). The predicted mean plasma p‐tau217 z‐scores for plasma are on the left y‐axis, the raw values are on the right y‐axis, and age in years is represented on the x‐axis. The observed data points are shown in gray. Bands represent 95% confidence intervals. Estimates are truncated to be within the age range of participants for a particular group. The predicted age‐related trajectories of plasma p‐tau217 are observed to be higher among APOE ε4 carriers, male participants, and those with a higher LIBRA index within the healthy control group (A, B, and C). Across the whole sample with known A status, changes in these trajectories were associated significantly with several factors in separate mixed‐effects models (D, E, and F). Specifically, APOE genotype influenced the rate of change, with the fastest decline seen in ε4/ε4 carriers, followed by ε3/ε4, ε2/ε4, ε3/ε3, and ε2/ε3, in that order. In addition, amyloid status (A+ vs A–) and amyloid Centiloid values also significantly impacted p‐tau217 trajectories. The effects varied significantly across different Centiloid ranges, with those  >21.6 showing the most pronounced changes, followed by the 14.1 to 21.6 range, and the least changes observed in values,14.1.

FIGURE 3

Interaction plot from mixed‐effects model of cognitive composite scores for plasma p‐tau217 (pg/mL). p‐tau217 <0.4, Low, 0.4 ≤ p‐tau217 ≤ 0.63, Intermediate, > 0.63 High risk groups. The lines represent the median plasma p‐tau217 values at baseline PACC3 (A), EF (B), Immediate memory (C), Delayed memory (D) and CDR‐SB (E) in the low, intermediate and high p‐tau217 risk groups described previously (Ashton et al.³). People with higher p‐tau217 values have higher rates of decline at older ages compared to people with lower p‐tau217 values.