Timing of Changes in Alzheimer's Disease Plasma Biomarkers as Assessed by Amyloid and Tau PET Clocks

Marta Milà‐Alomà; Duygu Tosun; Suzanne E Schindler; Isabella Hausle; Kellen K Petersen; Yan Li; Jeffrey L Dage; Lei Du‐Cuny; Ziad S Saad; Benjamin Saef; Gallen Triana‐Baltzer; David L Raunig; Janaky Coomaraswamy; Michael Baratta; Emily A Meyers; Yulia Mordashova; Carrie E Rubel; Kyle Ferber; Hartmuth Kolb; Nicholas J Ashton; Henrik Zetterberg; Erin G Rosenbaugh; Martin Sabandal; Leslie M Shaw; Anthony W Bannon; William Z Potter; for the Alzheimer's Disease Neuroimaging Initiative (ADNI); Foundation for the National Institutes of Health (FNIH) Biomarkers Consortium Plasma Aβ and Phosphorylated Tau as Predictors of Amyloid and Tau Positivity in Alzheimer's Disease Project Team

doi:10.1002/ana.27285

. 2025 Jun 20;98(3):508–523. doi: 10.1002/ana.27285

Timing of Changes in Alzheimer's Disease Plasma Biomarkers as Assessed by Amyloid and Tau PET Clocks

Marta Milà‐Alomà ^1,², Duygu Tosun ^1,^2,^✉, Suzanne E Schindler ³, Isabella Hausle ^1,², Kellen K Petersen ³, Yan Li ³, Jeffrey L Dage ^4,⁵, Lei Du‐Cuny ⁶, Ziad S Saad ⁷, Benjamin Saef ³, Gallen Triana‐Baltzer ⁷, David L Raunig ⁸, Janaky Coomaraswamy ⁸, Michael Baratta ⁸, Emily A Meyers ⁹, Yulia Mordashova ⁶, Carrie E Rubel ¹⁰, Kyle Ferber ¹⁰, Hartmuth Kolb ¹¹, Nicholas J Ashton ¹², Henrik Zetterberg ^12,^13,^14,^15,^16,¹⁷, Erin G Rosenbaugh ¹⁸, Martin Sabandal ¹⁸, Leslie M Shaw ¹⁹, Anthony W Bannon ²⁰, William Z Potter ^†; for the Alzheimer's Disease Neuroimaging Initiative (ADNI)^‡; Foundation for the National Institutes of Health (FNIH) Biomarkers Consortium Plasma Aβ and Phosphorylated Tau as Predictors of Amyloid and Tau Positivity in Alzheimer's Disease Project Team

^¹Northern California Institute for Research and Education, San Francisco, CA, USA

^²Department of Radiology and Biomedical Imaging, University of California San Francisco, San Francisco, CA, USA

^³Department of Neurology, Washington University in St. Louis, St. Louis, MO, USA

^⁴Department of Neurology, Indiana University School of Medicine, Indianapolis, IN, USA

^⁵Stark Neurosciences Research Institute, Indiana University School of Medicine, Indianapolis, IN, USA

^⁶AbbVie, Ludwigshafen am Rhein, Rheinland‐Pfalz, Germany

^⁷Neuroscience Biomarkers, Johnson and Johnson Innovative Medicine, San Diego, CA, USA

^⁸Takeda Pharmaceutical Company Ltd. Cambridge, MA, USA

^⁹Alzheimer's Association, Chicago, IL, USA

^¹⁰Biogen, Cambridge, MA, USA

^¹¹Enigma Biomedical Group, San Diego, CA, USA

^¹²Institute of Neuroscience and Physiology, Department of Psychiatry and Neurochemistry, The Sahlgrenska Academy at University of Gothenburg, Mölndal, Sweden

^¹³Clinical Neurochemistry Laboratory, Sahlgrenska University Hospital, Mölndal, Sweden

^¹⁴UK Dementia Research Institute Fluid Biomarkers Laboratory, UK DRI at UCL, London, UK

^¹⁵Department of Neurodegenerative Disease, UCL Queen Square Institute of Neurology, London, UK

^¹⁶Hong Kong Center for Neurodegenerative Diseases, Clear Water Bay, Hong Kong, China

^¹⁷Wisconsin Alzheimer's Disease Research Center, University of Wisconsin School of Medicine and Public Health, University of Wisconsin‐Madison, Madison, WI, USA

^¹⁸The Foundation for the National Institutes of Health, North Bethesda, MD, USA

^¹⁹Department of Pathology and Laboratory Medicine, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA

^²⁰AbbVie, North Chicago, IL, USA

Address correspondence to Dr Tosun, Department of Radiology and Biomedical Imaging, University of California San Francisco, 505 Parnassus Avenue, San Francisco, CA 94143, USA. E‐mail: duygu.tosun@ucsf.edu

^†

Highly qualified expert.

^‡

Alzheimer's Disease Neuroimaging Initiative (ADNI): Data used in preparation of this article were obtained from the Alzheimer's Disease Neuroimaging Initiative (ADNI) database (adni.loni.usc.edu). As such, the investigators within the ADNI contributed to the design and implementation of ADNI and/or provided data but did not participate in analysis or writing of this report. A complete listing of ADNI investigators can be found at: http://adni.loni.usc.edu/wp-content/uploads/how_to_apply/ADNI_Acknowledgement_List.pdf.

^✉

Corresponding author.

PMCID: PMC12335434 PMID: 40539416

Abstract

Objective

The objective of this study was to evaluate the timing of change of Alzheimer's disease (AD) plasma biomarkers (Aβ42/Aβ40, p‐tau217, p‐tau181, GFAP, and NfL) from six different assay platforms, alongside established AD biomarkers, using amyloid and tau positron emission tomography (PET)‐based AD progression timelines.

Methods

Data from the Alzheimer's Disease Neuroimaging Initiative (ADNI), including 784 individuals with longitudinal amyloid PET and 359 with longitudinal tau PET, were analyzed to estimate the age at amyloid and tau PET positivity, respectively. Longitudinal plasma biomarker measurements were available from 190 individuals with an estimated amyloid PET positivity age and from 70 individuals with an estimated tau PET positivity age. In a subset of 17 clinical progressors, age at tau PET positivity strongly predicted symptom onset, allowing for estimation of symptom onset age. Biomarker trajectories based on time from amyloid or tau PET positivity or symptom onset were modelled using Generalized Additive Mixed models. Time intervals of significant biomarker change and the earliest timepoints at which biomarkers exceeded predefined abnormality thresholds were identified.

Results

All plasma biomarkers except NfL became abnormal prior to established thresholds for amyloid and tau PET positivity. Plasma Aβ42/Aβ40 became abnormal very early in both amyloid PET and tau PET timelines, while plasma GFAP became abnormal early in the tau PET timeline. Plasma Aβ42/Aβ40 levels plateaued, whereas plasma p‐tau217, p‐tau181, GFAP, and NfL levels increased throughout the modeled disease progression. Some variations in the timing of these changes were observed across different biomarker assays.

Interpretation

These findings suggest that the plasma Aβ42/Aβ40 may be useful in identifying individuals with very low levels of amyloid pathology, whereas p‐tau, GFAP, and NfL may be useful in staging disease progression. ANN NEUROL 2025;98:508–523

Alzheimer's disease (AD) is neuropathologically defined by the extracellular accumulation of amyloid plaques primarily comprised of amyloid‐β (Aβ) peptide, and intraneuronal aggregation of neurofibrillary tangles, which consist of hyperphosphorylated tau protein. ¹ These pathological hallmarks progressively accumulate in individuals during the asymptomatic preclinical phase of AD, often leading to the symptomatic phase characterized by mild cognitive impairment (MCI) and dementia. ² Biomarkers used to detect AD pathology in vivo include positron emission tomography (PET) scans and the measurement of protein levels in the cerebrospinal fluid (CSF) or blood. ³ Blood‐based biomarkers (BBMs) offer several potential advantages over PET and CSF biomarkers, including reduced burden and greater accessibility of testing, potentially accelerating recruitment for clinical trials and expand access to AD diagnostics and treatments. ⁴ , ⁵ , ⁶ Available BBMs include Aβ42/Aβ40, ⁷ p‐tau181, ⁸ p‐tau217, ⁹ the ratio of p‐tau217 to non‐phosphorylated tau (%p‐tau217), ¹⁰ neurofilament light (NfL), ¹¹ and glial fibrillary acidic protein (GFAP). ¹² Some BBMs, such as Aβ42/Aβ40 and p‐tau217, become abnormal during the preclinical phase of AD and are currently used in certain prevention clinical trials to identify individuals at risk of cognitive decline. ¹³ , ¹⁴ , ¹⁵ Additionally, the observed responsiveness of some BBMs to therapeutic interventions suggests their potential utility in monitoring treatment efficacy. ¹⁶ , ¹⁷

This study investigated when BBMs became abnormal compared with established thresholds for amyloid or tau PET positivity. It is important to clarify that the “timing of abnormality” in this context refers to when biomarker values begin to deviate from a defined normal range, which precedes the point at which biomarkers reach established positivity thresholds. For example, the established thresholds for amyloid or tau PET positivity are conservatively set to maximize specificity, such that individuals may exhibit detectable amyloid or tau pathology changes on PET for many years before reaching established positivity thresholds. ¹⁸ , ¹⁹ Amyloid or tau PET positivity were used as reference points in this study due to their widespread use across research and clinical settings, providing a common framework for comparison. Understanding when BBM levels change relative to amyloid or tau PET positivity informs our understanding of AD biology, may inform clinical trial design, and could guide the clinical interpretation of BBM results.

To examine the timing of change in BBMs, we utilized a “biological clock” approach, leveraging the well‐documented, sequential accumulation of brain pathology as measured by PET. ¹⁹ , ²⁰ , ²¹ Using longitudinal amyloid and tau PET data from the ADNI cohort, we estimated the age at which participants reached amyloid PET positivity, tau PET positivity, and the onset of AD symptoms. This approach allowed us to align individual's biomarker trajectories based on estimated years from these key events, enabling a more accurate assessment of the relative timing of biomarker changes. We then evaluated the trajectories of BBMs, along with established (non‐plasma) AD biomarkers as a function of estimated years from amyloid PET positivity, tau PET positivity, and AD symptom onset, as well as determined the relative timing of biomarker changes.

Materials and Methods

Study Participants

The current study utilized data from the ADNI database (adni.loni.usc.edu), which aims to develop and validate biomarkers for AD clinical trials. ²² , ²³ Details of the ADNI cohort can be found in Supplementary Methods Data S1. Amyloid and tau PET clock models were developed using all available ADNI longitudinal ¹⁸F‐florbetapir (FBP) amyloid PET (n = 784) and ¹⁸F‐flortaucipir (FTP) tau PET (n = 359) data from the ADNI study (Supplementary Table S1). The timing of BBM change was modeled using data from 292 individuals within the FNIH Biomarkers Consortium “trajectories” study dataset, hereafter referred to as the “main study cohort.” This cohort comprised cognitively unimpaired (CU) and cognitively impaired (CI) ADNI participants with longitudinal BBM measurements (n = 292), amyloid PET scans (n = 292), and tau PET scans (n = 204) with all participants having BBM measurements taken at three different timepoints. ²⁴ For modeling the timing of change in established (non‐plasma) AD biomarkers including cortical thickness, CSF p‐tau181/Aβ42, amyloid PET, mesial‐temporal and temporo‐parietal tau PET, and the Clinical Dementia Rating Sum of Boxes (CDR‐SB), the analysis utilized the broader ADNI dataset; see Supplementary Figure S1 for details on the study design and sample sizes for these analyses. Participants were required to have at least one positivite amyloid PET or tau PET scan, based on a predefined threshold, to be included in the estimation of age at amyloid or tau PET positivity, as descrbied below.

Plasma Biomarker Measurements

Plasma samples were collected, processed, and stored according to the standardized ADNI protocols. ²⁵ Analyses were performed using the following assays: C2N Diagnostics PrecivityAD2 (C2N), Fujirebio Diagnostics Lumipulse (Fujirebio), ALZpath Quanterix (ALZpath), Janssen LucentAD Quanterix (Janssen), and Roche Diagnostics NeuroToolKit (Roche) assays; a subset (84%) additionally underwent analysis with the Quanterix Neurology 4‐Plex (Quanterix) assays. ²⁴

CSF Biomarker Measurements and APOE Genotyping

CSF Aβ42, total tau, and p‐tau181 were measured using the Elecsys® β‐amyloid (1–42), Elecsys Total‐Tau, and Elecsys Phospho‐Tau (181P) immunoassays, respectively, on a Cobas e 601 analyzer (Roche Diagnostics International Ltd, Rotkreuz, Switzerland). APOE genotyping was performed as part of the standard ADNI protocol.

Amyloid PET, Tau PET, and MRI Measures

Amyloid PET imaging with FBP was conducted at each ADNI site adhering to standardized protocols. ²⁶ Global standardized uptake value ratios (SUVRs) were calculated by averaging across cortical regions (frontal, cingulate, parietal, and lateral temporal) defined by FreeSurfer version 7.1. SUVRs were normalized to a composite reference region (whole cerebellum, brainstem/pons, and subcortical eroded white matter) as recommended in the ADNI for longitudinal analyses. ²⁷ Centiloid values were derived using the transformations provided by the ADNI PET Core. ²⁸ A Gaussian mixture model (GMM) with two components was applied to cross‐sectional FBP‐PET data from the entire ADNI study (n = 1,347), to determine a positivity threshold. The threshold was set at the mean plus two standard deviations of the first component of the GMM (low amyloid burden), resulting in 0.78 SUVR. This threshold aligns with the ADNI study standard and approximately corresponds to 26 Centiloid (ADNI_UCBerkeley_AmyloidPET_Methods_v2_2023‐06‐29.pdf; Supplementary Fig S2).

Tau PET imaging with FTP was conducted at each ADNI site following standardized protocols. SUVRs were calculated for a mesial‐temporal (entorhinal, parahippocampus, and amygdala) meta‐region of interest (ROI) defined by FreeSurfer version 7.1. SUVRs were also calculated for a temporo‐parietal meta‐ROI (banks of superior temporal, cuneus, inferior–superior parietal, inferior‐middle‐superior temporal, istmuscingulate, lateral occipital, lingal, posterior cingulate, precuneus, and supramarginal). SUVRs were normalized to the inferior cerebellar grey matter reference region, using the SUIT template. ²⁹ The same GMM method used for amyloid PET was applied to mesial‐temporal meta‐ROI data to establish a threshold for early tau PET positivity, yielding an SUVR threshold of 1.41 (see Supplementary Fig S2).

Structural brain magnetic resonance imaging (MRI) data (3D T1‐weighted images at 3T magnetic field) were acquired at each site. The full cortex of each brain was parcellated into 41 bilateral ROIs using the Desikan‐Killiany‐Tourville atlas in FreeSurfer (version 5.1 for ADNI‐GO/2, and version 6.0 for ADNI‐3). A composite meta‐ROI cortical thickness measurement was calculated, encompassing entorhinal, fusiform, parahippocampal, middle temporal, inferior temporal, and angular gyrus. ³⁰ To account for differences in image acquisition and processing, the meta‐ROI cortical thickness measure was harmonized using the ComBat‐GAM method, ³¹ , ³² adjusting for effects of normal aging and sex. All available ADNI data were utilized for unbiased modeling in the harmonization process (see Supplementary Methods Data S1).

Neuropsychological and Clinical Variables

Participants underwent comprehensive clinical assessments, including a detailed interview of a collateral source, a neurological examination of the participant, the Clinical Dementia Rating® (CDR®) and the CDR‐SB. ³³ Participants were classified based on their clinical diagnosis as CU, MCI, or dementia. For individuals diagnosed with MCI or dementia, clinicians assessed the likely etiology (AD or non‐AD) based on clinical presentation. ³⁴

Statistical Analyses

Creation of Amyloid and Tau PET Clocks

To construct amyloid and tau PET‐based progression timelines (referred to as “clocks”), we refined a previously established approach. ²⁰ Briefly, linear mixed‐effects models with random intercepts and slopes were used to estimate the rates of change in global amyloid or mesial‐temporal tau PET SUVRs over time. To account for nonlinear relationships between accumulation rates and pathology burden, we fitted generalized additive models (GAMs) with cubic splines. To ensure robust rate estimation, clock models were restricted to the SUVR range where the variance of the GAM‐fitted values was below the 90th percentile of variance across all fitted values (0.62–1.11 SUVR for amyloid PET and 0.98–2.04 SUVR for tau PET; Supplementary Fig S3). Within this range, the time elapsed between 0.0001 SUVR unit increases was calculated by integrating the inverse of the modeled rates of change. Amyloid time and tau time were conceptualized as the estimated time since an individual reached the established positivity threshold for amyloid PET or mesial‐temporal tau PET, respectively. For individuals with at least one positive amyloid PET or tau PET scan, the age at amyloid or tau PET positivity (the estimated age at which they reached the positivity threshold) was calculated by subtracting the corresponding amyloid time or tau time from their age at the scan. For individuals with multiple scans, the average age at positivity across all scans was used (Supplementary Methods Data S1).

To validate these time estimates, we examined individuals who transitioned from a negative to positive amyloid PET or tau PET scan during their participation in the ADNI study and tested the correlations between the estimated and the actual time of transition (Supplementary Methods Data S1). The estimated years from amyloid or tau PET positivity were calculated as the difference between the age at an outcome measurement and the estimated age at amyloid or tau PET positivity, respectively.

Estimated Years from Symptom Onset

Symptom onset age was defined as the age at the first assessment with a CDR global score greater than zero and a clinical diagnosis of MCI or dementia due to AD. We first identified ADNI participants who clinically progressed from CU to MCI or dementia due to AD by their last assessment (clinical progressors) and had available amyloid or tau PET data. Participants were excluded if they developed a non‐AD neurodegenerative condition, remained CU at their last assessment, or had a negative amyloid PET at the time of clinical progression. Linear regression models were used to predict symptom onset age based on age at amyloid PET positivity, age at tau PET positivity, APOE ε4 status, and sex. The most parsimonious model was selected and applied to all participants. As a validation, we correlated the actual symptom onset age with the estimated symptom onset age in clinical progressors (Supplementary Methods Data S1). For subsequent analyses, the estimated years from symptom onset was calculated by subtracting the estimated age at symptom onset from the age at the outcome measure.

Timing of Change in AD Biomarkers Relative to the Amyloid Timeline, Tau Timeline, and Symptom Timeline

Generalized additive mixture models (GAMMs) with cubic spline basis functions and random intercepts were used to model each outcome measure (BBMs, CSF p‐tau181/Aβ42, amyloid PET, mesial‐temporal tau PET, temporo‐parietal tau PET, cortical thickness, and CDR‐SB) as a function of each estimated progression timeline (years from amyloid PET positivity, tau PET positivity, or symptom onset). Confidence intervals for the GAMM model estimates were calculated as the estimate ± (z value × standard error). Time periods of significant rate of change were identified by detecting time segments where the first derivative of the GAMMs significantly deviated from zero. Finite differencing with Taylor series expansions was used to approximate derivatives.

A reference group consisting of individuals who remained amyloid and tau PET negative and CU (CDR of 0 and clinical diagnosis of CU) throughout their participation in the ADNI study was selected. Confidence intervals for the mean of each outcome measure in the reference group were calculated as the mean ± (t‐value × standard error). The earliest time of abnormality for each outcome measure in each estimated timeline was defined as the first timepoint where the 95% confidence interval of the model estimate no longer overlapped with the 95% confidence interval of the overall mean value in the reference group. This approach prioritized sensitivity to detect early, subtle deviations from the reference group, rather than focusing solely on positivity thresholds. The median and 95% confidence intervals of the times of abnormality were calculated using 1,000 bootstrapped samples. Additionally, the timings of abnormality and of significant rate of change in amyloid and tau timelines were mapped to the corresponding amyloid or tau burden, expressed in Centiloid units or tau PET SUVR values, respectively.

As a sensitivity analysis, separate linear regression models were used to assess the associations between each outcome measure and age, sex, and APOE ε4 status within the reference group. For outcome measures with significant associations with these factors, raw values in the entire sample were adjusted for these factors using the regression models, and the timing of changes was re‐analyzed using the adjusted values. A separate sensitivity analysis was conducted by restricting the cohort to individuals with available plasma measures across all assays (n = 233), to evaluate the robustness of the results independent of sample size differences among the different outcome measures.

All statistical analyses and figure generation were performed using R software (R version 4.2.2).

Results

Characteristics of the Study Cohorts

In the main study cohort with BBMs, 190 participants had at least one positive amyloid PET scan, allowing for the estimation of their age at amyloid PET positivity; 70 participants had at least one positive tau PET scan, enabling estimation of their age at tau PET positivity. Forty‐eight participants overlapped between the two cohorts. Eighty individuals who remained amyloid and tau PET negative and CU throughout the study (CDR = 0 and clinical diagnosis of CU) served as the reference group.

In the combined cohort, the mean age was 73.9 ± 7.0 years, 147 (50.3%) were women, 121 (41.4%) were APOE ε4 carriers, 276 (94.5%) self‐identified as white. 112 (38.4%) were cognitively impaired (CDR > 0), and 130 (44.5%) were amyloid PET positive at their first plasma sample collection (see the Table 1). Compared to the reference group, individuals in both the cohort with amyloid PET positivity age estimates and the cohort with tau PET positivity age estimates exhibited a higher prevalence of APOE ε4 carriers and elevated levels of established AD biomarkers. Individuals in the cohort with tau PET positivity age estimates were significantly older, and individuals with amyloid PET positivity age estimates had significantly lower education levels than the reference group. Notably, only a small proportion of participants in either cohorts had a clinical diagnosis of dementia. Annual rates of change for all outcome measures in the main study cohort are provided in Supplementary Table S2. Plasma p‐tau217, p‐tau‐181, GFAP, and NfL measures exhibited significant annual rates of change in all groups, even within the reference group.

TABLE 1.

Characteristics of Participants in the Main Study Cohort

	Combined (n = 292)	Reference Group ^a (n = 80)	Cohort With Estimated Age at Amyloid PET Positivity ^b (n = 190)		Cohort With Estimated Age at Tau PET Positivity ^c (n = 70)
	Combined (n = 292)	Reference Group ^a (n = 80)	Values	p	Values	p
Demographics
Age, yr	73.9 (7.0)	72.7 (6.4)	74.2 (7.1)	0.08	75.0 (7.31)	0.04
Women, n (%)	147 (50.3)	36 (45.0)	100 (52.6)	0.31	37 (52.9)	0.43
APOE ε4 carriers, n (%)	121 (41.4)	17 (21.2)	97 (51.1)	< 0.0001	38 (54.3)	< 0.0001
Education attainment (Postgraduate), n (%)	128 (43.8)	43 (53.8)	73 (38.4)	0.03	31 (44.3)	0.32
Race (Black/White/Other), n (%)	10 (3.4)/276 (94.5)/6 (2.1)	3 (3.8)/75 (93.8)/2 (2.4)	6 (3.2)/182 (95.8)/2.0 (2.0)	0.51	1 (1.4)/66 (94.3)/3 (4.3)	0.42
Clinical/cognitive assessments
CDR 0/0.5/1, n (%)	180 (61.6)/108 (37.0)/4 (1.4)	80 (100)/‐/‐	89 (46.8)/97 (51.1)/4 (2.1)	< 0.0001	36 (51.4)/34 (48.6)/0 (0.0)	< 0.0001
CU/MCI/dementia, n (%)	179 (61.3)/105 (36.0)/8 (2.7)	80 (100.0)/‐/‐	87 (45.8)/95 (50.0)/8 (4.2)	< 0.0001	36 (51.4)/31 (44.3)/3 (4.3)	< 0.0001
CDR‐SB	0.6 (1.0)	0.0 (0.0)	0.9 (1.2)	< 0.0001	0.7 (0.9)	< 0.0001
AD biomarkers
CSF p‐tau181/Aβ42	0.03 (0.02)	0.01 (0.00)	0.0 (0.0)	< 0.0001	0.04 (0.02)	< 0.0001
Amyloid PET, SUVR	0.81 (0.14)	0.7 (0.0)	0.9 (0.1)	< 0.0001	0.8 (0.2)	< 0.0001
Amyloid PET‐positive, n (%)	130 (44.5)	0.0 (0.0)	127 (66.8)	<0.0001	38 (54.3)	<0.0001
Cortical thickness meta‐ROI, mm	2.81 (0.14)	2.8 (0.1)	2.8 (0.2)	0.02	2.8 (0.2)	0.04
Mesial‐temporal tau PET, SUVR	1.30 (0.27)	1.2 (0.1)	1.4 (0.3)	0.05	1.6 (0.3)	0.001
Temporo‐parietal tau PET, SUVR	1.19 (0.25)	1.1 (0.1)	1.1 (0.1)	0.99	1.4 (0.4)	0.14

Open in a new tab

Note: Characteristics of the study participants with longitudinal plasma biomarker measures at their first plasma assessment. Data are shown as mean and standard deviation or percentage (%), as appropriate. Continuous variables were compared using t test and categorical variables using Chi squared test. The p values refer to differences in relation to the reference group. Significant p values (p < 0.05) are shown in bold. n = 48 individuals had at least one positive amyloid PET scan and one positive tau PET scan throughout the study duration and therefore overlap between the two cohorts. Of note, the cohort with an estimated age at symptom onset is identical to the cohort with an estimated age at tau PET positivity (n = 80).

^{^a}

The reference group is comprised of individuals who were amyloid and tau PET‐negative and had a CDR = 0 and a clinical classification of cognitively unimpaired (CU) throughout the ADNI study duration.

^{^b}

The cohort with an estimated age at amyloid PET positivity comprises individuals with at least one positive amyloid PET scan throughout the study duration.

^{^c}

The cohort with an estimated age at tau PET positivity comprises individuals with at least one positive tau PET scan throughout the study duration.

AD = Alzheimer's disease; CDR = Clinical Dementia Rating; CDR‐SR = Clinical Dementia Rating Sum of Boxes; CSF = cerebrospinal fluid; CU = cognitively unimpaired; MCI = mild cognitive impairment; PET = positron emission tomography; ROI = region of interest; SUVR = standardized uptake value ratio.

Characteristics of the larger ADNI cohorts included in the analyses of established (non‐plasma) AD biomarkers and CDR‐SB (n = 635 with amyloid PET positivity age estimates, and n = 212 with tau PET positivity age estimates) are presented in Supplementary Table S3.

Creation of Amyloid and Tau PET Clocks

Among individuals who transitioned from negative to positive on amyloid or tau PET scans during their participation in the ADNI study, a strong correlation was observed between their estimated and actual ages at positivity (Spearman's ρ = 0.96, p < 0.0001 for n = 81 amyloid PET phenoconverters and Spearman's ρ = 0.96, p < 0.0001 for n = 35 tau PET phenoconverters; Supplementary Fig S4).

In the main study cohort, the mean estimated ages at amyloid PET positivity, tau PET positivity, and symptom onset were 66.6 ± 10.2, 73.7 ± 9.5, and 81.0 ± 6.1 years, respectively. Figure 1A and 1C illustrate the observed amyloid and tau PET SUVRs across different ages, demonstrating inter‐individual variability in the age at onset of amyloid and tau PET accumulation. Figure 1B and 1D present the modeled biomarker trajectories as a function of estimated time since amyloid or tau PET positivity, respectively, demonstrating a relatively consistent accumulation rate over time.

Amyloid and tau PET trajectories as a function of age and years from amyloid and tau PET positivity. Individual amyloid (A, B) and tau (C, D) PET longitudinal trajectories, where each line connects multiple observations from the same individual. Points and line segments are color‐coded according to the cognitive status assigned at each observation (blue for CDR = 0, orange for CDR = 0.5, and red for CDR ≥ 1). Triangles represent *APOE* ε4 carriers and circles represent *APOE* ε4 non‐carriers. One value with a ¹⁸F‐flortaucipir mesial‐temporal tau PET SUVR > 4 was excluded for visualization. CDR = Clinical Dementia Rating; PET = positron emission tomography; SUVR = standardized uptake value ratio. [Color figure can be viewed at www.annalsofneurology.org]

Estimation of the Age at Symptom Onset

Within the complete ADNI dataset, the age at tau PET positivity was a very strong predictor of the age at symptom onset in clinical progressors with tau PET positivity (n = 17), with an adjusted R ² of 0.86 (p < 0.0001). In a larger set of clinical progressors with amyloid PET positivity (n = 39), the age at amyloid PET positivity was a much weaker predictor of the age at symptom onset with an adjusted R ² of 0.38 (p < 0.0001). In models predicting the age at symptom onset that included both the age at tau PET positivity and the age at amyloid PET positivity, only the age at tau PET positivity remained a significant predictor. Therefore, the final model for estimating the age at symptom onset included only the age at tau PET positivity (Supplementary Fig S5). Notably, neither APOE ε4 status nor sex significantly influenced the estimation of age at symptom onset.

AD Biomarker Trajectories as a Function of the Amyloid Timeline, Tau Timeline, and Symptom Timeline

Figure 2 illustrates the modeled trajectories of established AD biomarkers and CDR‐SB across the three estimated AD progression timelines. Figure 3 presents the trajectories for BBMs previously reported to exhibit stronger associations with amyloid PET and tau PET. ²⁴ Trajectories for the remaining BBMs are provided in Supplementary Figures S6 and S7. Supplementary Figures S8 to S10 depict the trajectories of all outcome measures as a function of chronological age, global amyloid PET SUVR, and mesial‐temporal tau PET SUVR.

Established AD biomarkers and CDR‐SB as a function of estimated AD progression timelines. GAMMs with cubic spline basis and random intercepts were used to model established AD biomarkers and CDR‐SB as a function of years from amyloid PET positivity (A), years from tau PET positivity (B), and years from symptom onset (C). Dashed horizontal lines indicate the mean of the reference group, with shaded areas representing the 95% confidence interval. Solid vertical lines mark the time of abnormality compared to the reference group. Time periods with a significant rate of change are indicated with thicker lines. Individual trajectories are shown, where each line connects multiple observations from the same individual. Points and line segments are color‐coded according to the cognitive status assigned at each observation (blue for CDR = 0, orange for CDR = 0.5, and red for CDR ≥ 1). Triangles represent *APOE* ε4 carriers and circles represent *APOE* ε4 non‐carriers. One value with ¹⁸F‐flortaucipir mesial‐temporal tau PET SUVR > 4 and six values with ¹⁸F‐flortaucipir temporo‐parietal tau PET SUVR > 4 were excluded for visualization. AD = Alzheimer's disease; CDR = Clinical Dementia Rating; CDR‐SB = Clinical Dementia Rating Sum of Boxes; GAMMs = Generalized Additive Mixed Models; PET = positron emission tomography; SUVR = standardized uptake value ratio. [Color figure can be viewed at www.annalsofneurology.org]

AD plasma biomarkers as a function of estimated AD progression timelines. GAMMs with cubic spline basis and random intercepts were used to model plasma biomarkers as a function of years from amyloid PET positivity (A), years from tau PET positivity (B), and years from symptom onset (C). Dashed horizontal lines indicate the mean of the reference group, with shaded areas representing the 95% confidence interval. Solid vertical lines mark the time of abnormality compared to the reference group. Time periods with a significant rate of change are indicated with thicker lines. Individual trajectories are shown, where each line connects multiple observations from the same individual. Points and line segments are color‐coded according to the cognitive status assigned at each observation (blue for CDR = 0, orange for CDR = 0.5, and red for CDR ≥ 1). Triangles represent *APOE* ε4 carriers and circles represent *APOE* ε4 non‐carriers. Outlier values were excluded for visualization: one for Roche Aβ42/Aβ40, two for Fujirebio Aβ42/Aβ40, one for Janssen p‐tau217, two for C2N p‐tau217, two for ALZpath p‐tau217, three for Fujirebio p‐tau217, three for Quanterix p‐tau181, three for Roche GFAP, two for Quanterix GFAP, one for Roche NfL, and one for Quanterix NfL. AD = Alzheimer's disease; CDR = Clinical Dementia Rating; GAMMs = Generalized Additive Mixed Models; PET = positron emission tomography. [Color figure can be viewed at www.annalsofneurology.org]

Timing of AD Biomarker Changes Relative to the Amyloid Timeline

For the amyloid timeline, the estimated time of reaching the amyloid PET threshold was set as time zero and data from 190 individuals were included in the models. In Figures 2 and 3, thicker segments of the fitted lines indicate time periods with significant biomarker rates of change. CSF p‐tau181/Aβ42, amyloid PET, mesial‐temporal tau PET, cortical thickness, and CDR‐SB exhibited significant rates of change throughout the entire modeled amyloid timeline. In contrast, increases in temporo‐parietal tau PET started 5.8 years after the amyloid PET positivity threshold (see Fig 2A, and Supplementary Table S4). Plasma p‐tau217, p‐tau181, GFAP, and NfL consistently increased throughout the entire amyloid timeline. Conversely, plasma Aβ42/Aβ40 initially decreased (−7.9 to −7.1 years prior to the threshold for amyloid PET positivity) but plateaued at +11.5 to +13.0 years after amyloid PET positivity, with slight variations depending on the specific assay used (see Fig 3A, Supplementary Figs S6A and S7A, and Supplementary Table S4).

As expected, amyloid PET levels exhibited detectable abnormalities before reaching the positivity threshold, at −5.9 years, closely followed by CSF p‐tau181/Aβ42 at −5.2 years. Abnormalities in mesial‐temporal and temporo‐parietal tau PET SUVR and CDR‐SB emerged around the same time as amyloid PET positivity, whereas cortical thickness exhibited abnormalities after amyloid PET positivity at +2.9 years. All BBMs except for plasma NfL reached abnormal levels before the amyloid PET positivity threshold. Plasma Aβ42/Aβ40, as measured by C2N and Roche assays, showed the earliest abnormalities, emerging prior to amyloid PET positivity at −7.9 years, followed by Fujirebio Aβ42/Aβ40 at −5.8 years. In contrast, plasma Aβ42/Aβ40 as measured by Quanterix assay exhibited abnormalities significantly later, based on confidence intervals comparison, at −3.4 years. Abnormalities in plasma C2N %p‐tau217 followed at −4.4 years, along with GFAP, Janssen, and ALZpath p‐tau217, and Roche p‐tau181, all ranging from −3.9 to −2.9 years. Fujirebio p‐tau217 reached abnormal levels at −1.5 years, whereas Quanterix p‐tau181 and C2N p‐tau217 did not become abnormal until around the estimated time of amyloid PET positivity. Plasma NfL became abnormal last at +9.2 to +10.5 years after amyloid PET positivity, depending on the assay used (see Fig 4A, and Supplementary Table S5).

Timing for AD biomarker abnormality. Estimated years from amyloid PET positivity (A), tau PET positivity (B), or symptom onset (C) where each outcome measure becomes abnormal compared to the reference group. A and B also show the timing for AD biomarker abnormality in Centiloid units or mesial‐temporal tau PET SUVR values, respectively. Points depict the median and error bars depict the 95% confidence interval. Dashed vertical lines at time = 0 represent the time at amyloid PET positivity (¹⁸F‐florbetapir global cortical amyloid PET SUVR > 0.78) (A), tau PET positivity (¹⁸F‐flortaucipir mesial‐temporal tau PET SUVR > 1.41) (B) or symptom onset (C). Established AD biomarkers and CDR‐SB are shaded in grey. AD = Alzheimer's disease; CDR‐SR = Clinical Dementia Rating Sum of Boxes; PET = positron emission tomography; SUVR = standardized uptake value ratio. [Color figure can be viewed at www.annalsofneurology.org]

The timing of biomarker changes, expressed as corresponding amyloid burden in Centiloid units, is also provided in Figure 4A and in Supplementary Table S6.

Timing of AD Biomarker Changes Relative to the Tau Timeline

For the tau timeline, with the estimated time of reaching the tau PET threshold set as time zero, data from 70 individuals were included in the models. Amyloid PET, mesial‐temporal and temporo‐parietal tau PET, and CSF p‐tau181/Aβ42 consistently increased throughout the entire tau timeline. In contrast, CDR‐SB began to increase prior to tau PET positivity at −6.3 years; however, this earlier timing might be influenced by the very low variance in CDR‐SB within the reference group. Cortical thickness began to decrease shortly after tau PET positivity at +0.5 years and continued to decline until the upper limit of the timeline at +18.1 years (see Fig 2B, and Supplementary Table S4). The temporal dynamics of BBMs across the tau timeline varied depending on the assay used. C2N and Roche Aβ42/Aβ40 exhibited consistent decreases throughout the timeline. In contrast, Quanterix Aβ42/Aβ40 plateaued at +6 years, and Fujirebio Aβ42/Aβ40 exhibited a significant decrease only from −0.8 to +4.4 years. All p‐tau217 measures demonstrated significant increases throughout the entire tau timeline, except for C2N p‐tau217, which began to rise significantly at −1.0 years. Roche GFAP and Roche p‐tau181 exhibited consistent increases across the entire tau timeline, whereas Quanterix p‐tau181 increases were significant only until +3.6 years. Notably, Quanterix plasma GFAP and NfL measures did not exhibit significant rates of change at any point during the tau timeline (see Fig 3B, Supplementary Figs S6B and S7B, and Supplementary Table S4).

Abnormalities in mesial‐temporal and temporo‐parietal tau PET were detected before the positivity threshold at −4.7 and −0.6 years, respectively. Amyloid PET and CSF p‐tau181/Aβ42 also became abnormal prior to tau PET positivity at −7.8 and −7.4 years, respectively. Cortical thickness became abnormal later, but still before tau PET positivity at −2.8 years. CDR‐SB exhibited abnormalities before medial‐temporal tau PET positivity at −5.7 years, a finding that may again reflect limited variability in CDR‐SB in the reference group. Plasma GFAP exhibited the earliest estimated deviations from the reference group in the tau timeline, occurring at −11.5 years. However, due to overlapping confidence intervals, the timing of GFAP change was not significantly different from the estimated timing of change for C2N, Fujirebio, and Roche Aβ42/Aβ40 (ranging from −9.3 to −7.1 years), Roche and Quanterix p‐tau181 (−6.7 or − 5.4 years, respectively) and C2N p‐tau217 (−4.7 years). ALZpath, Janssen, and Fujirebio p‐tau217, and C2N %p‐tau217 showed abnormalities slightly later at −5.4 to −4.9 years. Finally, plasma NfL did not show abnormal levels until after tau PET positivity at +5.3 or +15.3 years for Roche and Quanterix, respectively (Fig 4B, and see Supplementary Table S5).

The timing of these biomarker changes expressed as the corresponding mesial‐temporal tau PET SUVR values is also provided in Figure 4B and in Supplementary Table S6.

Timing of AD Biomarker Changes Relative to the Symptom Timeline

For the symptom timeline, the estimated time of developing symptoms was set as time zero and data from 70 individuals were included in the models (the same individuals as for the tau PET timeline). Global amyloid PET, mesial‐temporal and temporo‐parietal tau PET, and CSF p‐tau181/Aβ42 exhibited sustained changes throughout the entire modeled symptom timeline, whereas the rate of change in CDR‐SB and cortical thickness became significant prior to symptom onset at −8.6 and −6.2 years, respectively, and persisted throughout the timeline (see Fig 2C and Supplementary Table S5).

Rates of change for plasma Aβ42/Aβ40 varied across assays, with Quanterix plasma Aβ42/Aβ40 showing a significant rate of change over the longest interval (−16.7 to −0.8 years) and no plasma Aβ42/Aβ40 assay demonstrating a significant rate of change after estimated symptom onset. In contrast, plasma p‐tau217, as well as Quanterix p‐tau181, GFAP, and NfL, demonstrated steady increases across the entire symptom timeline (see Fig 3C, Supplementary Figs S6C and S7C, and Supplementary Table S4).

Amyloid PET and CSF p‐tau181/Aβ42 showed abnormalities prior to the estimated symptom onset at −14.4 years. Abnormalities in mesial‐temporal and temporo‐parietal tau PET were detected shortly after, at −11.9 and −11.7 years, respectively. Abnormalities in CDR‐SB and in cortical thickness also appeared prior to symptom onset at −6.2 and −2.1 years, respectively. Roche, C2N, and Fujirebio plasma Aβ42/Aβ40 were the earliest BBMs to exhibit abnormalities, appearing at up to −14.4 years. In contrast, Quanterix plasma Aβ42/Aβ40 exhibited a significantly later deviation from the reference group, estimated at −8.4 years, as indicated by non‐overlapping confidence intervals. Plasma GFAP, p‐tau217, and p‐tau181 all reached abnormality prior to symptom onset within a range of −11.7 to −9.4 years. Plasma NfL changed last, at −4.3 or −2.9 years depending on the assay (Fig 4C, and see Supplementary Table S5).

Sensitivity Analyses

Neither sex nor APOE ε4 status had a statistically significant effect on any of the outcome measures within the reference group. In contrast, mesial‐temporal and temporo‐parietal tau PET, and all plasma p‐tau181, GFAP, NfL, and p‐tau217 measures, except for C2N p‐tau217 and %p‐tau217, demonstrated significant associations with age. When assessed using age‐corrected values, the relative order of these biomarkers across the different timelines, as well as their dynamic changes, closely resembled that observed with non‐corrected data (Supplementary Tables S7 and S8).

Finally, the timing of changes in all outcome measures was assessed in the subset of individuals with available plasma measures across all assays (n = 233), and the results were similar to those in the full study cohorts (Supplementary Tables S9 and S10).

Discussion

This study aimed to evaluate the timing of changes in BBMs relative to estimated AD progression timelines derived from amyloid PET and tau PET clocks. Our key findings indicate that plasma Aβ42/Aβ40 and GFAP demonstrated the earliest abnormalities. Importantly, plasma Aβ42/Aβ40 levels changed early but then plateaued, while plasma p‐tau217, p‐tau181, GFAP, and NfL levels increased steadily throughout the disease course. Notably, all plasma p‐tau measures became abnormal after amyloid PET first showed abnormality, yet prior to amyloid PET reaching the established threshold for positivity. As expected, plasma NfL was the last BBM to change. Overall, amyloid biomarkers changed first, followed by biofluid tau biomarkers (potentially reflecting a response to amyloid pathology), tau PET, and then neurodegeneration markers. These findings align with previous literature on the sequence of BBMs abnormalities ³⁵ , ³⁶ , ³⁷ and underscore the sensitivity of BBMs in detecting early pathological changes in AD. These results contribute to a refined understanding of AD pathophysiology and the dynamics of key BBMs throughout AD progression, which are crucial for optimizing their use in clinical trials and clinical practice.

Several previous studies have evaluated AD biomarker trajectories, including BBMs, using amyloid PET, ¹⁴ , ³⁸ , ³⁹ or other models ³⁵ , ³⁶ , ⁴⁰ as proxies of disease progression. However, to our knowledge, only one prior study used a timescale derived from amyloid PET. ¹⁹ Importantly, the Li et al ¹⁹ study used an earlier threshold for amyloid PET positivity, corresponding to the “tipping point” where amyloid accumulation is thought to begin to occur at a consistent rate (termed amyloid onset). In contrast, this study used the time since reaching the established thresholds for amyloid PET positivity and tau PET positivity. Because these thresholds for amyloid and tau PET positivity are widely used despite their lack of sensitivity to early changes, we used them in this study as references for considering the timing of biomarker changes.

Despite different references used for “time zero” in this study and the Li et al study, the overall order of biomarker abnormality across AD progression was similar between the two studies. The current study further validated the amyloid PET clock approach in a larger, multicenter sample with extensive longitudinal data from individuals at various disease stages. By incorporating tau PET imaging and evaluating key AD BBMs using multiple assay platforms, this study expanded upon previous findings, enabling cross‐assay comparisons and enhancing the robustness of the findings. The inclusion of a tau timescale significantly improved the ability to capture the complexity of biomarker changes, as the tau clock captures processes that may occur independently of amyloid pathology, offering a more comprehensive view of disease progression.

Compared to a reference group of cognitively unimpaired individuals who remained amyloid and tau PET negative, plasma Aβ42/Aβ40 biomarkers reached abnormality as early as 7.9 years before the established threshold for amyloid PET positivity (~26 Centiloid), 9.3 years before tau PET positivity, and 14.4 years before symptom onset. These changes occurred concurrently with or even earlier than the first abnormalities detected by amyloid PET or CSF p‐tau181/Aβ42. These findings are consistent with prior studies showing changes in plasma Aβ42/Aβ40 before amyloid PET positivity is reached. ³⁵ , ⁴¹ , ⁴² , ⁴³ Plasma Aβ42/Aβ40 may therefore hold potential for selecting study participants with very early stages of amyloid accumulation for prevention trials.

Plasma GFAP changed shortly after Aβ42/Aβ40 in the amyloid timeline, concurrently with plasma p‐tau217 and p‐tau181 measures. In the tau timeline, plasma GFAP showed the earliest nominal change, although this was not significantly different from plasma Aβ42/Aβ40. The early rise in plasma GFAP might reflect astrocytic response to early tau pathology in amyloid‐positive individuals, ¹² , ⁴⁴ , ⁴⁵ or potentially non‐AD‐related processes. Similarly, the timing of changes in plasma p‐tau181 and p‐tau217 in the tau timeline, despite overlapping confidence intervals, could reflect differences in their AD‐specificity, with plasma p‐tau217 potentially being more specific. ⁴⁶ Importantly, the tau timeline was based on tau PET signal in mesial‐temporal regions, which may reflect both AD‐related and age‐related tauopathy. ⁴⁷ Whereas a neocortical tau timeline might provide a more AD‐specific measure of progression, the limited number of individuals with neocortical tau‐positivity (n = 40) precluded reliable modeling. We acknowledge this as a limitation in interpreting the tau PET clock and its impact on biomarker ordering. Nonetheless, despite the potential influence of non‐AD processes on mesial‐temporal tau PET, our study found that all plasma biomarkers except NfL became abnormal years before mesial‐temporal tau accumulation reached a positivity threshold. Our findings highlight the complexity of defining tau PET positivity and its relationship to plasma biomarkers.

Although generally consistent across different assay platforms, some variations were observed in the timing of BBM changes when measured by different assays. The Quanterix plasma Aβ42/Aβ40 assay demonstrated a delayed change compared to other platforms, potentially explaining its lower association with amyloid PET observed in our previous cross‐sectional study. ²⁴ C2N %p‐tau217 showed an earlier increase than p‐tau217 concentration measures, particularly in the amyloid timeline, underscoring its higher correlation with amyloid PET that could be due to lessened effects of medical comorbidities such as chronic kidney disease on this measure. ²⁴ , ⁴⁸ In addition, the temporal dynamics of plasma biomarkers varied by assay, particularly for plasma Aβ42/Aβ40 in the tau and symptom timelines. These observations likely reflect inherent variability in the sensitivity, specificity, the specific molecular species measured, and the overall variance of different assay platforms, especially when comparing assays of different types (eg, mass spectrometry or immunoassays). ⁷ , ²⁴ , ⁴⁹ These assay‐specific differences are crucial considerations when selecting biomarkers for clinical and research applications.

In addition to characterizing changes in BBM, another major finding of this study was that the estimated age at tau PET positivity was a stronger predictor of symptom onset than the age at amyloid PET positivity, with an R ² of 0.86. Although this finding aligns with the strong evidence linking tau pathology to cognitive decline, ⁵⁰ , ⁵¹ , ⁵² , ⁵³ previous studies have reported a stronger predictive value for amyloid onset age. ²⁰ , ²¹ Several factors may contribute to these differences. ADNI uses FBP as the amyloid PET tracer, which may have lower sensitivity and precision compared to other amyloid PET tracers such as PiB. ²⁰ , ⁵⁴ In addition, ADNI is a multicenter study with inherent heterogeneity in clinical diagnosis practices across different sites. Despite these factors, the predictive value of age at amyloid PET positivity in our model (R ² = 0.38) was comparable to that observed in autosomal‐dominant AD (ADAD) studies, where parental age at symptom onset demonstrates an R ² ranging from 0.38 to 0.56. ⁵⁵ , ⁵⁶ This is especially noteworthy considering that, unlike the relatively pure AD pathology observed in ADAD cases, sporadic AD is influenced by numerous factors, including comorbid neuropathologies, which can significantly contribute to the variability in cognitive symptoms.

This study has several limitations. The biological clock approach, while valuable for assessing biomarker changes across disease progression, is subject to limitations. Specifically, the interpretation of biomarker positivity is contingent upon the chosen threshold, the cohort demographic, and processing procedures. These factors should be carefully considered when evaluating the generalizability of our results to other populations or settings. In addition, whereas our results demonstrate consistent group‐level patterns of biomarker changes, the observed interindividual variability, especially beyond established biomarker positivity thresholds, adds variance to individual‐level predictions. The sample sizes for tau and symptom timeline analyses were smaller compared with the amyloid timeline analyses, and the sample size for BBM analyses was smaller compared with non‐plasma biomarkers and cognitive measures. However, sensitivity analyses indicated that these sample size limitations did not significantly impact the main results. Additionally, the model used to estimate age at symptom onset was based on a relatively small sample of 17 participants, although it demonstrated strong predictive performance. While this model evaluated for the effect of the major AD risk factors such as APOE ε4 status and sex on the age at symptom onset, other factors that may influence symptom onset, such as comorbidities, social determinants of health, and additional genetic variants, were not included. Furthermore, the study included a relatively low frequency of individuals with clinical diagnosis of dementia, limiting the ability to investigate biomarker trajectories in advanced stages of the disease. Finally, ⁵⁷ the present study primarily focused on retrospective data from a predominantly white and highly educated population. Therefore, further studies in other populations are necessary to confirm these findings.

In summary, this study comprehensively evaluated the timing of change in key AD BBMs across AD progression timelines using amyloid PET and tau PET clocks. The findings demonstrate that plasma Aβ42/Aβ40 exhibits the earliest abnormalities across the estimated AD progression timelines, although its levels plateaued around the estimated time of symptom onset. In contrast, plasma p‐tau217, p‐tau181, GFAP, and NfL started changing later but then increased over time, potentially allowing for disease staging. Importantly, these results highlight the significant variability in BBM dynamics across analytes and assay platforms, underscoring the importance of careful selection to specific biomarkers and assays tailored to particular research questions and clinical applications.

AUTHOR CONTRIBUTIONS

Marta Milà‐Alomà: Conceptualization; data curation; formal analysis; funding acquisition; investigation; methodology; project administration; writing – original draft; writing – review and editing. Duygu Tosun: Conceptualization; data curation; formal analysis; funding acquisition; investigation; methodology; project administration; resources; supervision; writing – original draft; writing – review and editing. Suzanne E. Schindler: Conceptualization; data curation; formal analysis; funding acquisition; investigation; methodology; resources; supervision; writing – original draft; writing – review and editing. Isabella Hausle: Data curation; formal analysis; writing – review and editing. Kellen K. Petersen: Data curation; formal analysis; writing – review and editing. Yan Li: Data curation; formal analysis; writing – review and editing. Jeffrey L. Dage: Data curation; formal analysis; writing – review and editing. Lei Du‐Cuny: Data curation; formal analysis; writing – review and editing. Ziad S. Saad: Data curation; formal analysis; writing – review and editing. Benjamin Saef: Data curation; formal analysis; writing – review and editing. Gallen Triana‐Baltzer: Data curation; formal analysis; writing – review and editing. David L. Raunig: Data curation; formal analysis; writing – review and editing. Janaky Coomaraswamy: Data curation; formal analysis; writing – review and editing. Michael Baratta: Data curation; formal analysis; writing – review and editing. Emily A. Meyers: Data curation; formal analysis; writing – review and editing. Yulia Mordashova: Data curation; formal analysis; writing – review and editing. Carrie E. Rubel: Data curation; formal analysis; writing – review and editing. Kyle Ferber: Data curation; formal analysis; writing – review and editing. Hartmuth Kolb: Data curation; formal analysis; writing – review and editing. Nicholas J. Ashton: Data curation; formal analysis; writing – review and editing. Henrik Zetterberg: Data curation; formal analysis; writing – review and editing. Erin G. Rosenbaugh: Data curation; formal analysis; project administration; writing – review and editing. Martin Sabandal: Data curation; formal analysis; project administration; writing – review and editing. Leslie M. Shaw: Conceptualization; data curation; formal analysis; funding acquisition; writing – review and editing. Anthony W. Bannon: Conceptualization; data curation; formal analysis; funding acquisition; writing – review and editing. William Z. Potter: Conceptualization; data curation; formal analysis; writing – original draft; writing – review and editing.

Potential conflicts of interests

H.K. has served on scientific advisory boards and/or as a consultant for ALZpath and Roche, and has given lectures sponsored by Fujirebio and Roche. J.L.D. has/is served/serving as a consultant or on advisory boards for AlzPath Inc., and Quanterix. J.L.D. has received research support from Fujirebio and Roche Diagnostics in the past 2 years. J.L.D. has stock or stock options in AlzPath Inc. Y.L. is the co‐inventor of the technology “Novel Tau isoforms to predict onset of symptoms and dementia in Alzheimer's disease” which is in the process of licensing by C2N. N.J.A. has received speaking fees from Quanterix. These conflicts involved companies that are developers of the assays used in the study. S.E.S. has served on scientific advisory boards on biomarker testing and education for Eisai and Novo Nordisk and has received speaking fees for presentations on biomarker testing from Eisai, Eli Lilly, and Novo Nordisk. M.M.A., D.T., I.H., K.K.P., L.D.C., Z.S.S., B.S., G.T.B., D.L.R., J.C., M.B., E.A.M., Y.M., C.E.R., K.F., H.K., E.G.R., M.S., L.M.S., A.W.B., and W.Z.P. have nothing to report. [Correction added on 12 July 2025, after first online publication: The following text “S.E.S. has served on scientific advisory boards……. Novo Nordisk.” has been added in this version.]

Supporting information

Data S1. Supporting Information.

ANA-98-508-s002.docx^{(11MB, docx)}

Data S2. Supporting Information.

ANA-98-508-s001.docx^{(46.7KB, docx)}

Acknowledgments

The results of the study represent results of the Foundation for the National Institutes of Health (FNIH) Biomarkers Consortium “Biomarkers Consortium, Plasma Aβ and Phosphorylated Tau as Predictors of Amyloid and Tau Positivity in Alzheimer's Disease” Project. The study was made possible through the scientific and financial support of industry, academic, patient advocacy, and governmental partners. We are grateful for the contributions of the following project team members: Iwona Dobler (Takeda), John Hsiao (NIA), Maria Quinton (AbbVie), Patricia Saletti (Alzheimer's Drug Discovery Foundation), Christopher Weber (Alzheimer's Association). Funding partners of the project include AbbVie Inc., Alzheimer's Association®, Diagnostics Accelerator at the Alzheimer's Drug Discovery Foundation, Biogen, Janssen Research & Development, LLC, and Takeda Pharmaceutical Company Limited. Private‐sector funding for the study was managed by the Foundation for the National Institutes of Health. We recognize C2N Diagnostics, Fujirebio Diagnostics with the Indiana University National Centralized Repository for Alzheimer's Disease and Related Dementias Biomarker Assay Laboratory (NCRAD‐BAL), Quanterix, and Roche Diagnostics with the University of Gothenburg for performing the plasma biomarker analysis in this study. The NCRAD‐BAL is supported by a cooperative agreement grant (U24 AG021886) awarded to NCRAD by the National Institute on Aging. Elecsys β‐amyloid (1–42) CSF, Elecsys Phospho‐Tau (181P) CSF, and Elecsys Total‐Tau CSF assays are approved for clinical use. COBAS and ELECSYS are trademarks of Roche. All other product names and trademarks are the property of their respective owners. The NeuroToolKit is a panel of exploratory prototype assays designed to robustly evaluate biomarkers associated with key pathologic events characteristic of AD and other neurological disorders, used for research purposes only and not approved for clinical use (Roche Diagnostics International Ltd, Rotkreuz, Switzerland). Finally, we would like to acknowledge ADNI for the plasma samples and ADNI participant data analyzed in this study. Data collection and sharing for the Alzheimer's Disease Neuroimaging Initiative (ADNI) is funded by the National Institute on Aging (National Institutes of Health Grant U19 AG024904). The grantee organization is the Northern California Institute for Research and Education. In the past, ADNI has also received funding from the National Institute of Biomedical Imaging and Bioengineering, the Canadian Institutes of Health Research, and private sector contributions through the Foundation for the National Institutes of Health (FNIH) including generous contributions from the following: AbbVie, Alzheimer's Association; Alzheimer's Drug Discovery Foundation; Araclon Biotech; BioClinica, Inc.; Biogen; Bristol‐Myers Squibb Company; CereSpir, Inc.; Cogstate; Eisai Inc.; Elan Pharmaceuticals, Inc.; Eli Lilly and Company; EuroImmun; F. Hoffmann‐La Roche Ltd and its affiliated company Genentech, Inc.; Fujirebio; GE Healthcare; IXICO Ltd.; Janssen Alzheimer Immunotherapy Research & Development, LLC.; Johnson & Johnson Pharmaceutical Research & Development LLC.; Lumosity; Lundbeck; Merck & Co., Inc.; Meso Scale Diagnostics, LLC.; NeuroRx Research; Neurotrack Technologies; Novartis Pharmaceuticals Corporation; Pfizer Inc.; Piramal Imaging; Servier; Takeda Pharmaceutical Company Limited; and Transition Therapeutics.

Data availability

Data from this study and the study methodology report may be accessed from the ADNI Laboratory of Neuro Imaging (LONI) database: adni.loni.usc.edu. The annotated code used for study analyses is provided in full (see https://github.com/cind/alz-timing-mila-aloma-et-al-2025).

References

1. Jack CR, Andrews JS, Beach TG, et al. Revised criteria for diagnosis and staging of Alzheimer's disease: Alzheimer's Association workgroup. Alzheimers Dement 2024;20:5143–5169. 10.1002/alz.13859. [DOI] [PMC free article] [PubMed] [Google Scholar]
2. McDade EM. Alzheimer Disease. Continuum (Minneap Minn) 2022;28:648–675. 10.1212/CON.0000000000001131. [DOI] [PubMed] [Google Scholar]
3. Schindler SE, Atri A. The role of cerebrospinal fluid and other biomarker modalities in the Alzheimer's disease diagnostic revolution. Nat Aging 2023;3:460–462. 10.1038/s43587-023-00400-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
4. Hansson O, Blennow K, Zetterberg H, Dage J. Blood biomarkers for Alzheimer's disease in clinical practice and trials. Nat Aging 2023;3:506–519. 10.1038/s43587-023-00403-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
5. Schindler SE, Li Y, Li M, et al. Using Alzheimer's disease blood tests to accelerate clinical trial enrollment. Alzheimers Dement 2023;19:1175–1183. 10.1002/alz.12754. [DOI] [PMC free article] [PubMed] [Google Scholar]
6. Palmqvist S, Tideman P, Mattsson‐Carlgren N, et al. Blood biomarkers to detect Alzheimer disease in primary care and secondary care. JAMA 2024;332:e2413855. 10.1001/jama.2024.13855. [DOI] [PMC free article] [PubMed] [Google Scholar]
7. Zicha S, Bateman RJ, Shaw LM, et al. Comparative analytical performance of multiple plasma Aβ42 and Aβ40 assays and their ability to predict positron emission tomography amyloid positivity. Alzheimers Dement 2023;19:956–966. 10.1002/alz.12697. [DOI] [PMC free article] [PubMed] [Google Scholar]
8. Karikari TK, Pascoal TA, Ashton NJ, et al. Blood phosphorylated tau 181 as a biomarker for Alzheimer's disease: a diagnostic performance and prediction modelling study using data from four prospective cohorts. Lancet Neurol 2020;19:422–433. 10.1016/S1474-4422(20)30071-5. [DOI] [PubMed] [Google Scholar]
9. Janelidze S, Bali D, Ashton NJ, et al. Head‐to‐head comparison of 10 plasma phospho‐tau assays in prodromal Alzheimer's disease. Brain 2023;146:1592–1601. 10.1093/brain/awac333. [DOI] [PMC free article] [PubMed] [Google Scholar]
10. Meyer MR, Kirmess KM, Eastwood S, et al. Clinical validation of the PrecivityAD2 blood test: a mass spectrometry‐based test with algorithm combining %p‐tau217 and Aβ42/40 ratio to identify presence of brain amyloid. Alzheimers Dement 2024;20:3179–3192. 10.1002/alz.13764. [DOI] [PMC free article] [PubMed] [Google Scholar]
11. Ashton NJ, Janelidze S, Al Khleifat A, et al. A multicentre validation study of the diagnostic value of plasma neurofilament light. Nat Commun 2021;12:3400. 10.1038/s41467-021-23620-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
12. Benedet AL, Milà‐Alomà M, Vrillon A, et al. Differences between plasma and cerebrospinal fluid glial fibrillary acidic protein levels across the Alzheimer disease continuum. JAMA Neurol 2021;78:1471–1483. 10.1001/jamaneurol.2021.3671. [DOI] [PMC free article] [PubMed] [Google Scholar]
13. Rissman RA, Langford O, Raman R, et al. Plasma Aβ42/Aβ40 and phospho‐tau217 concentration ratios increase the accuracy of amyloid PET classification in preclinical Alzheimer's disease. Alzheimers Dement 2024;20:1214–1224. 10.1002/alz.13542. [DOI] [PMC free article] [PubMed] [Google Scholar]
14. Milà‐Alomà M, Ashton NJ, Shekari M, et al. Plasma p‐tau231 and p‐tau217 as state markers of amyloid‐β pathology in preclinical Alzheimer's disease. Nat Med 2022;28:1797–1801. 10.1038/s41591-022-01925-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
15. Janelidze S, Barthélemy NR, Salvadó G, et al. Plasma phosphorylated tau 217 and Aβ42/40 to predict early brain aβ accumulation in people without cognitive impairment. JAMA Neurol 2024;81:e242619. 10.1001/jamaneurol.2024.2619. [DOI] [PMC free article] [PubMed] [Google Scholar]
16. Pontecorvo MJ, Lu M, Burnham SC, et al. Association of Donanemab Treatment with Exploratory Plasma Biomarkers in early symptomatic Alzheimer disease: a secondary analysis of the TRAILBLAZER‐ALZ randomized clinical trial. JAMA Neurol 2022;79:1250–1259. 10.1001/jamaneurol.2022.3392. [DOI] [PMC free article] [PubMed] [Google Scholar]
17. Hu Y, Cho M, Sachdev P, et al. Fluid biomarkers in the context of amyloid‐targeting disease‐modifying treatments in Alzheimer's disease. Med 2024;5:1206–1226. 10.1016/j.medj.2024.08.004. [DOI] [PubMed] [Google Scholar]
18. Insel PS, Ossenkoppele R, Gessert D, et al. Time to amyloid positivity and preclinical changes in brain metabolism, atrophy, and cognition: evidence for emerging amyloid pathology in Alzheimer's disease. Front Neurosci 2017;11:281. 10.3389/fnins.2017.00281. [DOI] [PMC free article] [PubMed] [Google Scholar]
19. Li Y, Yen D, Hendrix RD, et al. Timing of biomarker changes in sporadic Alzheimer's disease in estimated years from symptom onset. Ann Neurol 2024;95:951–965. 10.1002/ana.26891. [DOI] [PMC free article] [PubMed] [Google Scholar]
20. Schindler SE, Li Y, Buckles VD, et al. Predicting symptom onset in sporadic Alzheimer disease with amyloid PET. Neurology 2021;97:e1823–e1834. 10.1212/WNL.0000000000012775. [DOI] [PMC free article] [PubMed] [Google Scholar]
21. Betthauser TJ, Bilgel M, Koscik RL, et al. Multi‐method investigation of factors influencing amyloid onset and impairment in three cohorts. Brain 2022;145:4065–4079. 10.1093/brain/awac213. [DOI] [PMC free article] [PubMed] [Google Scholar]
22. Weiner MW, Veitch DP, Aisen PS, et al. The Alzheimer's disease neuroimaging initiative 3: continued innovation for clinical trial improvement. Alzheimers Dement 2017;13:561–571. 10.1016/j.jalz.2016.10.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
23. Weiner MW, Aisen PS, Jack CR, et al. The Alzheimer's disease neuroimaging initiative: progress report and future plans. Alzheimers Dement 2010;6:202–211.e7. 10.1016/j.jalz.2010.03.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
24. Schindler SE, Petersen KK, Saef B, et al. Head‐to‐head comparison of leading blood tests for Alzheimer's disease pathology. Alzheimers Dement 2024;20:8074–8096. 10.1002/alz.14315. [DOI] [PMC free article] [PubMed] [Google Scholar]
25. Kang JH, Korecka M, Figurski MJ, et al. The Alzheimer's disease neuroimaging initiative 2 biomarker Core: a review of progress and plans. Alzheimers Dement 2015;11:772–791. 10.1016/j.jalz.2015.05.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
26. Jagust WJ, Landau SM, Koeppe RA, et al. The Alzheimer's disease neuroimaging initiative 2 PET Core: 2015. Alzheimers Dement 2015;11:757–771. 10.1016/j.jalz.2015.05.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
27. Koeppe RA. Data analysis for amyloid PET imaging: Longitudinal studies, 2013.
28. Royse SK, Minhas DS, Lopresti BJ, et al. Validation of amyloid PET positivity thresholds in centiloids: a multisite PET study approach. Alzheimers Res Ther 2021;13:99. 10.1186/s13195-021-00836-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
29. Diedrichsen J. A spatially unbiased atlas template of the human cerebellum. NeuroImage 2006;33:127–138. 10.1016/j.neuroimage.2006.05.056. [DOI] [PubMed] [Google Scholar]
30. Jack CR, Wiste HJ, Weigand SD, et al. Defining imaging biomarker cut points for brain aging and Alzheimer's disease. Alzheimers Dement 2017;13:205–216. 10.1016/j.jalz.2016.08.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
31. Pomponio R, Erus G, Habes M, et al. Harmonization of large MRI datasets for the analysis of brain imaging patterns throughout the lifespan. NeuroImage 2020;208:116450. 10.1016/j.neuroimage.2019.116450. [DOI] [PMC free article] [PubMed] [Google Scholar]
32. Tassi E, Bianchi AM, Calesella F, et al. Assessment of ComBat harmonization performance on structural magnetic resonance imaging measurements. Hum Brain Mapp 2024;45:e70085. 10.1002/hbm.70085. [DOI] [PMC free article] [PubMed] [Google Scholar]
33. Morris JC, Blennow K, Ivanoiu A, et al. The clinical dementia rating (CDR): current version and scoring rules. Neurology 1993;43:2412–2414. 10.1212/wnl.0b013e318221ad47. [DOI] [PubMed] [Google Scholar]
34. Aisen P, Petersen R, Donohue M, et al. Clinical core of the Alzheimer's disease neuroimaging initiative: progress and plans. Alzheimers Dement 2010;6:239–246. 10.1016/j.jalz.2010.03.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
35. Bilgel M, An Y, Walker KA, et al. Longitudinal changes in Alzheimer's‐related plasma biomarkers and brain amyloid. Alzheimers Dement 2023;19:4335–4345. 10.1002/alz.13157. [DOI] [PMC free article] [PubMed] [Google Scholar]
36. Cogswell PM, Lundt ES, Therneau TM, et al. Modeling the temporal evolution of plasma p‐tau in relation to amyloid beta and tau PET. Alzheimers Dement 2024;20:1225–1238. 10.1002/alz.13539. [DOI] [PMC free article] [PubMed] [Google Scholar]
37. Janelidze S, Berron D, Smith R, et al. Associations of plasma Phospho‐Tau217 levels with tau positron emission tomography in early Alzheimer disease. JAMA Neurol 2021;78:149–156. 10.1001/jamaneurol.2020.4201. [DOI] [PMC free article] [PubMed] [Google Scholar]
38. Palmqvist S, Insel PS, Stomrud E, et al. Cerebrospinal fluid and plasma biomarker trajectories with increasing amyloid deposition in Alzheimer's disease. EMBO Mol Med 2019;11:e11170. 10.15252/emmm.201911170. [DOI] [PMC free article] [PubMed] [Google Scholar]
39. Mattsson‐Carlgren N, Andersson E, Janelidze S, et al. Aβ deposition is associated with increases in soluble and phosphorylated tau that precede a positive tau PET in Alzheimer's disease. Sci Adv 2020;6:eaaz2387. 10.1126/sciadv.aaz2387. [DOI] [PMC free article] [PubMed] [Google Scholar]
40. Roe CM, Ances BM, Head D, et al. Incident cognitive impairment: longitudinal changes in molecular, structural and cognitive biomarkers. Brain 2018;141:3233–3248. 10.1093/brain/awy244. [DOI] [PMC free article] [PubMed] [Google Scholar]
41. Rembach A, Faux NG, Watt AD, et al. Changes in plasma amyloid beta in a longitudinal study of aging and Alzheimer's disease. Alzheimers Dement 2014;10:53–61. 10.1016/J.JALZ.2012.12.006. [DOI] [PubMed] [Google Scholar]
42. Pereira JB, Janelidze S, Stomrud E, et al. Plasma markers predict changes in amyloid, tau, atrophy and cognition in non‐demented subjects. Brain 2021;144:2826–2836. 10.1093/brain/awab163. [DOI] [PMC free article] [PubMed] [Google Scholar]
43. Schindler SE, Bollinger JG, Ovod V, et al. High‐precision plasma β‐amyloid 42/40 predicts current and future brain amyloidosis. Neurology 2019;93:e1647–e1659. 10.1212/WNL.0000000000008081. [DOI] [PMC free article] [PubMed] [Google Scholar]
44. Pereira JB, Janelidze S, Smith R, et al. Plasma GFAP is an early marker of amyloid‐β but not tau pathology in Alzheimer's disease. Brain 2021;144:3505–3516. 10.1093/brain/awab223. [DOI] [PMC free article] [PubMed] [Google Scholar]
45. Bellaver B, Povala G, Ferreira PCL, et al. Astrocyte reactivity influences amyloid‐β effects on tau pathology in preclinical Alzheimer's disease. Nat Med 2023;29:1775–1781. 10.1038/s41591-023-02380-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
46. Therriault J, Vermeiren M, Servaes S, et al. Association of Phosphorylated tau Biomarkers with Amyloid Positron Emission Tomography vs tau positron emission tomography. JAMA Neurol 80:188–199. 10.1001/jamaneurol.2022.4485. [DOI] [PMC free article] [PubMed] [Google Scholar]
47. Crary JF, Trojanowski JQ, Schneider JA, et al. Primary age‐related tauopathy (PART): a common pathology associated with human aging. Acta Neuropathol 2014;128:755–766. 10.1007/s00401-014-1349-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
48. Janelidze S, Barthélemy NR, He Y, et al. Mitigating the associations of kidney dysfunction with blood biomarkers of Alzheimer disease by using phosphorylated tau to Total tau ratios. JAMA Neurol 2023;80:516–522. 10.1001/jamaneurol.2023.0199. [DOI] [PMC free article] [PubMed] [Google Scholar]
49. Janelidze S, Teunissen CE, Zetterberg H, et al. Head‐to‐Head comparison of 8 plasma amyloid‐β 42/40 assays in Alzheimer disease. JAMA Neurol 2021;78:1375–1382. 10.1001/jamaneurol.2021.3180. [DOI] [PMC free article] [PubMed] [Google Scholar]
50. Pontecorvo MJ, Devous MD, Kennedy I, et al. A multicentre longitudinal study of flortaucipir (18F) in normal ageing, mild cognitive impairment and Alzheimer's disease dementia. Brain 2019;142:1723–1735. 10.1093/brain/awz090. [DOI] [PMC free article] [PubMed] [Google Scholar]
51. Biel D, Brendel M, Rubinski A, et al. Tau‐PET and in vivo Braak‐staging as prognostic markers of future cognitive decline in cognitively normal to demented individuals. Alzheimers Res Ther 2021;13:137. 10.1186/s13195-021-00880-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
52. Strikwerda‐Brown C, Hobbs DA, Gonneaud J, et al. Association of Elevated Amyloid and tau Positron Emission Tomography Signal with Near‐Term Development of Alzheimer disease symptoms in older adults without cognitive impairment. JAMA Neurol 2022;79:975–985. 10.1001/jamaneurol.2022.2379. [DOI] [PMC free article] [PubMed] [Google Scholar]
53. Ossenkoppele R, Pichet Binette A, Groot C, et al. Amyloid and tau PET‐positive cognitively unimpaired individuals are at high risk for future cognitive decline. Nat Med 2022;28:2381–2387. 10.1038/s41591-022-02049-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
54. Su Y, Flores S, Wang G, et al. Comparison of Pittsburgh compound B and florbetapir in cross‐sectional and longitudinal studies. Alzheimers Dement (N Y) 2019;11:180–190. 10.1016/j.dadm.2018.12.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
55. Bateman RJ, Xiong C, Benzinger TLS, et al. Clinical and biomarker changes in dominantly inherited Alzheimer's disease. N Engl J Med 2012;367:795–804. 10.1056/NEJMoa1202753. [DOI] [PMC free article] [PubMed] [Google Scholar]
56. Ryman DC, Acosta‐Baena N, Aisen PS, et al. Symptom onset in autosomal dominant Alzheimer disease: a systematic review and meta‐analysis. Neurology 2014;83:253–260. 10.1212/WNL.0000000000000596. [DOI] [PMC free article] [PubMed] [Google Scholar]
57. Weiner MW, Veitch DP, Miller MJ, et al. Increasing participant diversity in AD research: plans for digital screening, blood testing, and a community‐engaged approach in the Alzheimer's disease neuroimaging initiative 4. Alzheimers Dement 2023;19:307–317. 10.1002/alz.12797. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Data S1. Supporting Information.

ANA-98-508-s002.docx^{(11MB, docx)}

Data S2. Supporting Information.

ANA-98-508-s001.docx^{(46.7KB, docx)}

Data Availability Statement

[ana27285-bib-0001] 1. Jack CR, Andrews JS, Beach TG, et al. Revised criteria for diagnosis and staging of Alzheimer's disease: Alzheimer's Association workgroup. Alzheimers Dement 2024;20:5143–5169. 10.1002/alz.13859. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0002] 2. McDade EM. Alzheimer Disease. Continuum (Minneap Minn) 2022;28:648–675. 10.1212/CON.0000000000001131. [DOI] [PubMed] [Google Scholar]

[ana27285-bib-0003] 3. Schindler SE, Atri A. The role of cerebrospinal fluid and other biomarker modalities in the Alzheimer's disease diagnostic revolution. Nat Aging 2023;3:460–462. 10.1038/s43587-023-00400-6. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0004] 4. Hansson O, Blennow K, Zetterberg H, Dage J. Blood biomarkers for Alzheimer's disease in clinical practice and trials. Nat Aging 2023;3:506–519. 10.1038/s43587-023-00403-3. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0005] 5. Schindler SE, Li Y, Li M, et al. Using Alzheimer's disease blood tests to accelerate clinical trial enrollment. Alzheimers Dement 2023;19:1175–1183. 10.1002/alz.12754. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0006] 6. Palmqvist S, Tideman P, Mattsson‐Carlgren N, et al. Blood biomarkers to detect Alzheimer disease in primary care and secondary care. JAMA 2024;332:e2413855. 10.1001/jama.2024.13855. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0007] 7. Zicha S, Bateman RJ, Shaw LM, et al. Comparative analytical performance of multiple plasma Aβ42 and Aβ40 assays and their ability to predict positron emission tomography amyloid positivity. Alzheimers Dement 2023;19:956–966. 10.1002/alz.12697. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0008] 8. Karikari TK, Pascoal TA, Ashton NJ, et al. Blood phosphorylated tau 181 as a biomarker for Alzheimer's disease: a diagnostic performance and prediction modelling study using data from four prospective cohorts. Lancet Neurol 2020;19:422–433. 10.1016/S1474-4422(20)30071-5. [DOI] [PubMed] [Google Scholar]

[ana27285-bib-0009] 9. Janelidze S, Bali D, Ashton NJ, et al. Head‐to‐head comparison of 10 plasma phospho‐tau assays in prodromal Alzheimer's disease. Brain 2023;146:1592–1601. 10.1093/brain/awac333. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0010] 10. Meyer MR, Kirmess KM, Eastwood S, et al. Clinical validation of the PrecivityAD2 blood test: a mass spectrometry‐based test with algorithm combining %p‐tau217 and Aβ42/40 ratio to identify presence of brain amyloid. Alzheimers Dement 2024;20:3179–3192. 10.1002/alz.13764. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0011] 11. Ashton NJ, Janelidze S, Al Khleifat A, et al. A multicentre validation study of the diagnostic value of plasma neurofilament light. Nat Commun 2021;12:3400. 10.1038/s41467-021-23620-z. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0012] 12. Benedet AL, Milà‐Alomà M, Vrillon A, et al. Differences between plasma and cerebrospinal fluid glial fibrillary acidic protein levels across the Alzheimer disease continuum. JAMA Neurol 2021;78:1471–1483. 10.1001/jamaneurol.2021.3671. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0013] 13. Rissman RA, Langford O, Raman R, et al. Plasma Aβ42/Aβ40 and phospho‐tau217 concentration ratios increase the accuracy of amyloid PET classification in preclinical Alzheimer's disease. Alzheimers Dement 2024;20:1214–1224. 10.1002/alz.13542. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0014] 14. Milà‐Alomà M, Ashton NJ, Shekari M, et al. Plasma p‐tau231 and p‐tau217 as state markers of amyloid‐β pathology in preclinical Alzheimer's disease. Nat Med 2022;28:1797–1801. 10.1038/s41591-022-01925-w. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0015] 15. Janelidze S, Barthélemy NR, Salvadó G, et al. Plasma phosphorylated tau 217 and Aβ42/40 to predict early brain aβ accumulation in people without cognitive impairment. JAMA Neurol 2024;81:e242619. 10.1001/jamaneurol.2024.2619. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0016] 16. Pontecorvo MJ, Lu M, Burnham SC, et al. Association of Donanemab Treatment with Exploratory Plasma Biomarkers in early symptomatic Alzheimer disease: a secondary analysis of the TRAILBLAZER‐ALZ randomized clinical trial. JAMA Neurol 2022;79:1250–1259. 10.1001/jamaneurol.2022.3392. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0017] 17. Hu Y, Cho M, Sachdev P, et al. Fluid biomarkers in the context of amyloid‐targeting disease‐modifying treatments in Alzheimer's disease. Med 2024;5:1206–1226. 10.1016/j.medj.2024.08.004. [DOI] [PubMed] [Google Scholar]

[ana27285-bib-0018] 18. Insel PS, Ossenkoppele R, Gessert D, et al. Time to amyloid positivity and preclinical changes in brain metabolism, atrophy, and cognition: evidence for emerging amyloid pathology in Alzheimer's disease. Front Neurosci 2017;11:281. 10.3389/fnins.2017.00281. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0019] 19. Li Y, Yen D, Hendrix RD, et al. Timing of biomarker changes in sporadic Alzheimer's disease in estimated years from symptom onset. Ann Neurol 2024;95:951–965. 10.1002/ana.26891. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0020] 20. Schindler SE, Li Y, Buckles VD, et al. Predicting symptom onset in sporadic Alzheimer disease with amyloid PET. Neurology 2021;97:e1823–e1834. 10.1212/WNL.0000000000012775. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0021] 21. Betthauser TJ, Bilgel M, Koscik RL, et al. Multi‐method investigation of factors influencing amyloid onset and impairment in three cohorts. Brain 2022;145:4065–4079. 10.1093/brain/awac213. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0022] 22. Weiner MW, Veitch DP, Aisen PS, et al. The Alzheimer's disease neuroimaging initiative 3: continued innovation for clinical trial improvement. Alzheimers Dement 2017;13:561–571. 10.1016/j.jalz.2016.10.006. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0023] 23. Weiner MW, Aisen PS, Jack CR, et al. The Alzheimer's disease neuroimaging initiative: progress report and future plans. Alzheimers Dement 2010;6:202–211.e7. 10.1016/j.jalz.2010.03.007. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0024] 24. Schindler SE, Petersen KK, Saef B, et al. Head‐to‐head comparison of leading blood tests for Alzheimer's disease pathology. Alzheimers Dement 2024;20:8074–8096. 10.1002/alz.14315. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0025] 25. Kang JH, Korecka M, Figurski MJ, et al. The Alzheimer's disease neuroimaging initiative 2 biomarker Core: a review of progress and plans. Alzheimers Dement 2015;11:772–791. 10.1016/j.jalz.2015.05.003. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0026] 26. Jagust WJ, Landau SM, Koeppe RA, et al. The Alzheimer's disease neuroimaging initiative 2 PET Core: 2015. Alzheimers Dement 2015;11:757–771. 10.1016/j.jalz.2015.05.001. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0027] 27. Koeppe RA. Data analysis for amyloid PET imaging: Longitudinal studies, 2013.

[ana27285-bib-0028] 28. Royse SK, Minhas DS, Lopresti BJ, et al. Validation of amyloid PET positivity thresholds in centiloids: a multisite PET study approach. Alzheimers Res Ther 2021;13:99. 10.1186/s13195-021-00836-1. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0029] 29. Diedrichsen J. A spatially unbiased atlas template of the human cerebellum. NeuroImage 2006;33:127–138. 10.1016/j.neuroimage.2006.05.056. [DOI] [PubMed] [Google Scholar]

[ana27285-bib-0030] 30. Jack CR, Wiste HJ, Weigand SD, et al. Defining imaging biomarker cut points for brain aging and Alzheimer's disease. Alzheimers Dement 2017;13:205–216. 10.1016/j.jalz.2016.08.005. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0031] 31. Pomponio R, Erus G, Habes M, et al. Harmonization of large MRI datasets for the analysis of brain imaging patterns throughout the lifespan. NeuroImage 2020;208:116450. 10.1016/j.neuroimage.2019.116450. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0032] 32. Tassi E, Bianchi AM, Calesella F, et al. Assessment of ComBat harmonization performance on structural magnetic resonance imaging measurements. Hum Brain Mapp 2024;45:e70085. 10.1002/hbm.70085. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0033] 33. Morris JC, Blennow K, Ivanoiu A, et al. The clinical dementia rating (CDR): current version and scoring rules. Neurology 1993;43:2412–2414. 10.1212/wnl.0b013e318221ad47. [DOI] [PubMed] [Google Scholar]

[ana27285-bib-0034] 34. Aisen P, Petersen R, Donohue M, et al. Clinical core of the Alzheimer's disease neuroimaging initiative: progress and plans. Alzheimers Dement 2010;6:239–246. 10.1016/j.jalz.2010.03.006. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0035] 35. Bilgel M, An Y, Walker KA, et al. Longitudinal changes in Alzheimer's‐related plasma biomarkers and brain amyloid. Alzheimers Dement 2023;19:4335–4345. 10.1002/alz.13157. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0036] 36. Cogswell PM, Lundt ES, Therneau TM, et al. Modeling the temporal evolution of plasma p‐tau in relation to amyloid beta and tau PET. Alzheimers Dement 2024;20:1225–1238. 10.1002/alz.13539. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0037] 37. Janelidze S, Berron D, Smith R, et al. Associations of plasma Phospho‐Tau217 levels with tau positron emission tomography in early Alzheimer disease. JAMA Neurol 2021;78:149–156. 10.1001/jamaneurol.2020.4201. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0038] 38. Palmqvist S, Insel PS, Stomrud E, et al. Cerebrospinal fluid and plasma biomarker trajectories with increasing amyloid deposition in Alzheimer's disease. EMBO Mol Med 2019;11:e11170. 10.15252/emmm.201911170. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0039] 39. Mattsson‐Carlgren N, Andersson E, Janelidze S, et al. Aβ deposition is associated with increases in soluble and phosphorylated tau that precede a positive tau PET in Alzheimer's disease. Sci Adv 2020;6:eaaz2387. 10.1126/sciadv.aaz2387. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0040] 40. Roe CM, Ances BM, Head D, et al. Incident cognitive impairment: longitudinal changes in molecular, structural and cognitive biomarkers. Brain 2018;141:3233–3248. 10.1093/brain/awy244. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0041] 41. Rembach A, Faux NG, Watt AD, et al. Changes in plasma amyloid beta in a longitudinal study of aging and Alzheimer's disease. Alzheimers Dement 2014;10:53–61. 10.1016/J.JALZ.2012.12.006. [DOI] [PubMed] [Google Scholar]

[ana27285-bib-0042] 42. Pereira JB, Janelidze S, Stomrud E, et al. Plasma markers predict changes in amyloid, tau, atrophy and cognition in non‐demented subjects. Brain 2021;144:2826–2836. 10.1093/brain/awab163. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0043] 43. Schindler SE, Bollinger JG, Ovod V, et al. High‐precision plasma β‐amyloid 42/40 predicts current and future brain amyloidosis. Neurology 2019;93:e1647–e1659. 10.1212/WNL.0000000000008081. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0044] 44. Pereira JB, Janelidze S, Smith R, et al. Plasma GFAP is an early marker of amyloid‐β but not tau pathology in Alzheimer's disease. Brain 2021;144:3505–3516. 10.1093/brain/awab223. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0045] 45. Bellaver B, Povala G, Ferreira PCL, et al. Astrocyte reactivity influences amyloid‐β effects on tau pathology in preclinical Alzheimer's disease. Nat Med 2023;29:1775–1781. 10.1038/s41591-023-02380-x. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0046] 46. Therriault J, Vermeiren M, Servaes S, et al. Association of Phosphorylated tau Biomarkers with Amyloid Positron Emission Tomography vs tau positron emission tomography. JAMA Neurol 80:188–199. 10.1001/jamaneurol.2022.4485. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0047] 47. Crary JF, Trojanowski JQ, Schneider JA, et al. Primary age‐related tauopathy (PART): a common pathology associated with human aging. Acta Neuropathol 2014;128:755–766. 10.1007/s00401-014-1349-0. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0048] 48. Janelidze S, Barthélemy NR, He Y, et al. Mitigating the associations of kidney dysfunction with blood biomarkers of Alzheimer disease by using phosphorylated tau to Total tau ratios. JAMA Neurol 2023;80:516–522. 10.1001/jamaneurol.2023.0199. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0049] 49. Janelidze S, Teunissen CE, Zetterberg H, et al. Head‐to‐Head comparison of 8 plasma amyloid‐β 42/40 assays in Alzheimer disease. JAMA Neurol 2021;78:1375–1382. 10.1001/jamaneurol.2021.3180. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0050] 50. Pontecorvo MJ, Devous MD, Kennedy I, et al. A multicentre longitudinal study of flortaucipir (18F) in normal ageing, mild cognitive impairment and Alzheimer's disease dementia. Brain 2019;142:1723–1735. 10.1093/brain/awz090. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0051] 51. Biel D, Brendel M, Rubinski A, et al. Tau‐PET and in vivo Braak‐staging as prognostic markers of future cognitive decline in cognitively normal to demented individuals. Alzheimers Res Ther 2021;13:137. 10.1186/s13195-021-00880-x. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0052] 52. Strikwerda‐Brown C, Hobbs DA, Gonneaud J, et al. Association of Elevated Amyloid and tau Positron Emission Tomography Signal with Near‐Term Development of Alzheimer disease symptoms in older adults without cognitive impairment. JAMA Neurol 2022;79:975–985. 10.1001/jamaneurol.2022.2379. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0053] 53. Ossenkoppele R, Pichet Binette A, Groot C, et al. Amyloid and tau PET‐positive cognitively unimpaired individuals are at high risk for future cognitive decline. Nat Med 2022;28:2381–2387. 10.1038/s41591-022-02049-x. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0054] 54. Su Y, Flores S, Wang G, et al. Comparison of Pittsburgh compound B and florbetapir in cross‐sectional and longitudinal studies. Alzheimers Dement (N Y) 2019;11:180–190. 10.1016/j.dadm.2018.12.008. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0055] 55. Bateman RJ, Xiong C, Benzinger TLS, et al. Clinical and biomarker changes in dominantly inherited Alzheimer's disease. N Engl J Med 2012;367:795–804. 10.1056/NEJMoa1202753. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0056] 56. Ryman DC, Acosta‐Baena N, Aisen PS, et al. Symptom onset in autosomal dominant Alzheimer disease: a systematic review and meta‐analysis. Neurology 2014;83:253–260. 10.1212/WNL.0000000000000596. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ana27285-bib-0057] 57. Weiner MW, Veitch DP, Miller MJ, et al. Increasing participant diversity in AD research: plans for digital screening, blood testing, and a community‐engaged approach in the Alzheimer's disease neuroimaging initiative 4. Alzheimers Dement 2023;19:307–317. 10.1002/alz.12797. [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Timing of Changes in Alzheimer's Disease Plasma Biomarkers as Assessed by Amyloid and Tau PET Clocks

Marta Milà‐Alomà, PhD

Duygu Tosun, PhD

Suzanne E Schindler, MD, PhD

Isabella Hausle, MS

Kellen K Petersen, PhD

Yan Li, PhD

Jeffrey L Dage, PhD

Lei Du‐Cuny, PhD

Ziad S Saad, PhD

Benjamin Saef, MS

Gallen Triana‐Baltzer, PhD

David L Raunig, PhD

Janaky Coomaraswamy, PhD

Michael Baratta, BA, MCAHPM

Emily A Meyers, PhD

Yulia Mordashova, MS

Carrie E Rubel, PhD

Kyle Ferber, PhD

Hartmuth Kolb, PhD

Nicholas J Ashton, PhD

Henrik Zetterberg, MD, PhD

Erin G Rosenbaugh, PhD

Martin Sabandal, PhD

Leslie M Shaw, PhD

Anthony W Bannon, PhD

William Z Potter, MD, PhD

Abstract

Objective

Methods

Results

Interpretation

Materials and Methods

Study Participants

Plasma Biomarker Measurements

CSF Biomarker Measurements and APOE Genotyping

Amyloid PET, Tau PET, and MRI Measures

Neuropsychological and Clinical Variables

Statistical Analyses

Creation of Amyloid and Tau PET Clocks

Estimated Years from Symptom Onset

Timing of Change in AD Biomarkers Relative to the Amyloid Timeline, Tau Timeline, and Symptom Timeline

Results

Characteristics of the Study Cohorts

TABLE 1.

Creation of Amyloid and Tau PET Clocks

FIGURE 1.

Estimation of the Age at Symptom Onset

AD Biomarker Trajectories as a Function of the Amyloid Timeline, Tau Timeline, and Symptom Timeline

FIGURE 2.

FIGURE 3.

Timing of AD Biomarker Changes Relative to the Amyloid Timeline

FIGURE 4.

Timing of AD Biomarker Changes Relative to the Tau Timeline

Timing of AD Biomarker Changes Relative to the Symptom Timeline

Sensitivity Analyses

Discussion

AUTHOR CONTRIBUTIONS

Potential conflicts of interests

Supporting information

Acknowledgments

Data availability

References

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases