Hamid R. Sohrabi

Background
The agreement between plasma Aβ42/40 and Aβ positron emission tomography (PET) is approximately 75 %, with ∼85 % of discrepancies due to positive plasma but negative PET results. It is unclear whether this reflects Aβ changes in plasma before PET-detectable.

Objectives
To assess the influence of Aβ42/40 positivity on risk of progression to Aβ PET positivity, and feasibility of using plasma Aβ42/40 tests to enrich a primary prevention trial.

Design
A prospective longitudinal cohort study.

Setting
Participants of Australian Imaging, Biomarkers and Lifestyle study (AIBL), Alzheimer’s Disease Neuroimaging Initiative (ADNI), and Open Access Series of Imaging Studies 3 (OASIS3).

Participants
507 cognitively unimpaired adults at baseline, with a baseline Aβ PET < 20 Centiloid (CL) and available longitudinal Aβ PET data.

Measurements
Baseline Aβ PET and plasma Aβ42/40 measurement by mass-spectrometry, followed by 1–6 additional Aβ PET scans every 1.5–3 years. Those < 5 CL were classified as PET- and 5–20 CL as PETLow. Plasma -/+ was defined using the Aβ42/40 Youden’s Index threshold (0.119), corresponding to Aβ PET status.

Results
At baseline, 283 were Plasma-/PET-, 97 Plasma+/PET-, 76 Plasma-/PETLow, and 51 Plasma+/PETLow. Among Plasma+/PET- individuals, 19 % progressed to PET+ (>20 CL), indicating a higher risk of progression, compared to Plasma-/PET- (HR: 3.90 [90 % CI: 2.00–7.61], p < 0.001). This elevated risk remained significant after matching the groups’ baseline CL (3.43 [1.43–8.26], p = 0.010), or adjustment for age, sex, APOE ε4 and baseline CL (2.48 [1.22 - 5.07], p = 0.013). Plasma+/PET- individuals accumulated Aβ ∼8 times faster (1.14 CL/year) than Plasma-/PET- (0.15 CL/year, p < 0.001). Plasma+/PET- progressors became PET+ two years earlier than Plasma-/PET- progressors. Among the Plasma+/PETLow individuals, 67 % progressed to PET+. Their progression was faster and earlier than in the Plasma-/PETLow group (HR: 20.82 [11.28 - 38.42], p < 0.001 vs. 6.67 [3.51 - 12.65], p < 0.001; reference: Plasma-/PET-), largely driven by higher baseline CL in the Plasma+ group. In a primary prevention paradigm targeting high-risk PETLow individuals, pre-screening with Aβ42/40 blood test reduced the number of PET scans by 49 %, compared to a PET-only strategy.

Conclusions
Cognitively unimpaired individuals with abnormal Aβ42/40 are at increased risk for future Aβ PET positivity. In the 5–20 CL subgroup, baseline CL is the main driver of this risk. Combining blood-based pre-screening with PET imaging may help efficiently enrich primary prevention trials.

Introduction
Around 90% of aged care residents with dementia have hearing and/or vision impairment. These impairments frequently go unsupported, exacerbating the impact of dementia and leading to loss of independence, social isolation, anxiety, and depression. We worked with residents, family and care-workers to co-design a sensory support intervention that included a sensory champion, care-worker training, integration of hearing/vision support into daily tasks, visual reminders and hearing/vision support stations to store devices (e.g. glasses and hearing aids).

Objectives
To evaluate implementation of the co-designed hearing and vision intervention using the Theoretical Framework of Acceptability (TFA).

Method
The intervention was evaluated via a TFA (5-Likert scale) questionnaire and semi-structured interview data with care-workers in two sites. Quantitative data were analysed using descriptive statistics and qualitative data via thematic analysis guided by TFA constructs (affective attitude, perceived effectiveness, intervention coherence, self-efficacy, burden, ethicality, and opportunity costs.)

Results
Thirty-five care-workers completed the questionnaire and 13 were interviewed. Questionnaire responses indicated high acceptability (M=4.71, SD=0.46), and perceived effectiveness. Participants reported that the intervention increased provision of hearing/vison support (M=4.46, SD=0.56) and improved residents’ outcomes (M=4.66, SD=0.48). Ratings where withing the ‘neural’ range for burden (M=3.17, SD=1.25) and opportunity costs (M=3.23, SD=1.19). Thematic analysis highlighted positive attitudes, and perceived effectiveness of the intervention including greater staff awareness of hearing/vision needs, improved communication, and better management of sensory aids. Care-workers reported they had developed skills in hearing aid maintenance and cleaning, and had integrated hearing/vision support into daily care routines. Participants found that the intervention was easy to implement and suggested periodic training to sustain knowledge.

Conclusion
The intervention was perceived by care workers as being highly acceptable and effective, with care workers integrating hearing/vision support for residents into their daily routines. Ongoing training could enhance long-term sustainability of hearing/vision support in aged care settings.

Aim: To identify the optimal Aβ PET/MMSE combination for predicting progression from mild cognitive impairment (MCI) to Alzheimer's disease (AD) dementia.

Methods: 686 MCI participants with Clinical Dementia Rating (CDR) score of 0.5 were followed for up to 7 years. Harmonized data from AIBL (n = 166) and ADNI (n = 520) were analysed using Cox proportional-hazards models, adjusted for age, sex, and APOE4 status, with the event of interest being progression to mild dementia due to AD (CDR = 1 and Aβ>20 CL). Optimal thresholds for MMSE (27) and Aβ (44 CL) were selected to maximize hazard ratios (HR) at 3 years. Combination of both MMSE and Aβ cut-offs provided four groups 1) high-cognition, low-Aβ (reference group), 2) low-cognition, low-Aβ, 3) high-cognition, high-Aβ, and 4) low-cognition, high-Aβ. Cognitive trajectories over time were modelled by harmonized Preclinical Alzheimer's Cognitive Composite (PACC) scores using linear mixed models, stratified by combined groups and adjusted for age, sex, education and APOE status.

Results: Both 44 CL and MMSE = 27 thresholds showed comparable hazard ratios (HR = 1.50 and 1.49, respectively; Figure 1). However, the MCI high-cognition group had higher progression risk (risk probability (RP) = 1-survival probability = 0.17±0.05) than MCI low-Aβ (RP = 0.01±0.01). Combining both cutoffs improved risk stratification: 75 out of 135 MCI low-cognition, high-Aβ progressed to AD within 3 years (30% survival probability, HR=2.9), while only 1 of 308 of the MCI high-cognition low-Aβ progressed to AD (RP = 0.04±0.03). The Linear mixed-effect models indicated that the low-cognition high-Aβ MCIs group showed the fastest decline with an effect size of 0.75, compared to the high-cognition, low-Aβ group.

Conclusion: Cognitive performance alone is not a sufficient substitute for Aβ cutoffs in predicting MCI-to-AD progression. However, binary classification based on Aβ can be improved by combining Aβ with MMSE to further stratify the risk of progression, providing greater prognostic information on an individual level and aiding in the design of clinical trials and therapeutic interventions for prodromal AD.
[Display Omitted]

Background
Dementia is currently the second leading cause of death in Australia and the seventh leading cause of death worldwide. Diagnosis of Alzheimer's disease (AD), the major cause of dementia, is difficult and time consuming. Current clinical imaging technologies are costly to use for widespread early screening of AD and have limited availability. In contrast, the retina is unique, where blood vessels and neural tissue can be viewed and imaged non-invasively and relatively inexpensively. As part of the central nervous system, the retina exhibits similarities to the brain and can display indicators of various neurological disorders, including AD. We aimed to image the retina and analyse its spectral features to develop a robust machine learning (ML) classification model that distinguishes between brain amyloid-beta (Aβ) positive and negative individuals.

Method
Sixty-eight consenting volunteers with varying levels of brain Aβ protein underwent non-invasive imaging using a hyperspectral retinal camera and illumination of wavelengths from 450 to 905 nm. Multiple retina features from the central and superior views were selected and analysed to identify their variability among individuals with different brain amyloid loads. Eight ML models were evaluated for their performance in predicting brain Aβ levels using the retina images and systemic factors like age, gender and apolipoprotein E (APOE) genotype.

Result
The retinal reflectance spectra in the 405–585 nm wavelengths exhibited a significant difference in individuals with increasing brain amyloid. The retinal features in the superior view showed higher inter-subject variability. Our comparison of eight different ML algorithms revealed that the Multi-Layer Perceptron (MLP) model exhibited superior classification performance, achieving an accuracy of 0.86, precision of 0.88, recall of 0.82, F1-score of 0.85, and area under curve (AUC) of 0.90.

Conclusion
This study indicates that there are spectral variations of retinal features associated with brain amyloid loads. It also demonstrates the feasibility of the retinal hyperspectral imaging (rHSI) technique as a potential non-invasive, inexpensive screening method to identify individuals in the preclinical phase of AD.

Short-chain fatty acids (SCFAs) produced by gut microbial fermentation influence host metabolism and neuroinflammatory processes implicated in Alzheimer’s disease (AD). However, the relationship between fecal SCFAs, microbial taxa, and cerebral amyloid-β (Aβ) burden in cognitively unimpaired individuals remains unclear. Fecal SCFAs were quantified using GC-MS, and microbial species were profiled by shotgun metagenomics in 87 participants. Associations between SCFAs, demographics, APOE ε4 status, and Aβ burden were tested using nonparametric statistics and multivariable regression. Microbial–SCFA links were evaluated using Spearman correlations and multivariate ordinations, with mediation analysis exploring potential indirect pathways. Acetate was the predominant SCFA and demonstrated the most robust microbial associations. Higher acetate concentrations were positively associated with Bacteroides ovatus and Faecalibacterium prausnitzii, whereas lower acetate levels were linked to species such as Bifidobacterium animalis and Lachnoclostridium scindens. Stratified analyses indicated that individuals with elevated Aβ burden exhibited more pronounced species–SCFA relationships, including a notable association between Bacteroides thetaiotaomicron and butyrate. Multivariate ordination further identified a significant overall coupling between SCFA profiles and microbial community structure. Mediation analysis suggested that an Oscillospiraceae species may represent a potential intermediary linking valerate concentrations with Aβ status. SCFA concentrations were not strongly influenced by demographic or genetic factors, but specific species demonstrated robust associations with acetate levels. Distinct SCFA–microbial interaction patterns in Aβ High individuals suggest subtle early gut microbial alterations linked to amyloid burden. These findings highlight the potential role of SCFA-related microbial pathways in preclinical AD.

Background: CSF and blood soluble TREM2 (sTREM2) levels have been found to increase at early stage of Alzheimer's disease (AD). The relationships between sTREM2, AD-related biomarkers, and other neuroinflammation biomarkers remain unclear. Moreover, the impact of rare variants in TREM2 gene (R47H/R62H), which are associated with increased risk of AD, on plasma sTREM2 has not been elucidated.

Objective: Investigate the association of plasma sTREM2 levels with brain amyloid-beta (A beta) load and AD-related blood biomarkers, i.e., phosphorylated tau (pTau)-181, pTau-231, GFAP, NFL, and other neuroinflammation and peripheral inflammation markers in cognitively normal (CN) older adults at risk of AD (CN A beta+) compared to CN A beta-, including the effect of AD-linked TREM2 rare variants.

Methods: Plasma sTREM2 concentrations were measured by MesoScale Discovery (MSD) assay from the KARVIAH cohort. Participants underwent cognitive tests and PET amyloid imaging. Genetic data and blood biomarkers were included for correlation analysis. Associations with plasma sTREM2 were investigated upon stratification by PET-A beta load SUVR ((CN A beta- (n = 65) and CN A beta+ (n = 35)) as the main analysis. A subgroup analysis based on the TREM2 R47H and R62H genotype was conducted as exploratory analysis.

Results: Plasma sTREM2 positively correlated with plasma pTau181, and pTau231 in CN A beta+ group. Plasma sTREM2 positively correlated with serum microglial kynurenine pathway metabolites. Plasma sTREM2 and brain A beta load were higher in R47H TREM2 carriers compared to non-carriers.

Conclusions: Our findings suggest plasma sTREM2 relates to downstream tau processes in amyloid-positive individuals, providing novel insights into the roles of peripheral TREM2 signaling that reflects microglial activity in early AD neuropathological development.

Background
This study aims to determine the optimal threshold for Aβ PET to identify individuals with mild cognitive impairment (MCI) who are at high risk of progressing to Alzheimer's disease (AD). Additionally, it assesses whether combining β‐amyloid with Mini‐Mental State Examination (MMSE) performance can enhance risk stratification in MCI which can guide clinical decision‐making regarding early therapeutic interventions.

Methods
We included 686 MCI participants with CDR 0.5 from two cohorts, followed for up to 7 years. Harmonized data from AIBL (N=166) and ADNI (N=520) were analysed using Cox proportional‐hazards models, adjusted for sex, and APOE4 status, with the event of interest being progression to mild dementia due to AD (detected by CDR = 1 and CL>20). Optimal thresholds for MMSE (27) and Aβ (44 CL) were selected to maximize hazard ratios (HR) at 3 years, categorizing participants into low‐risk and high‐risk groups based on cognitive performance and Aβ load. Note that the MMSE score was selected as it is frequently used in clinical practice and in trials.

Results
Both thresholds showed comparable hazard ratios (Figure 1). However, the MCI high‐cognition group had a significantly higher risk of progressing to AD (measured with risk probability (RP)=1‐survival probability; RP=0.08±0.05) than MCI low‐Aβ (RP=0.01±0.01). Combining both cutoffs improved risk stratification: 51 out of 135 MCI low‐cognition, high‐Aβ progressed to AD within 3 years (50% survival probability, HR=2.00), while only 1 of 308 of the MCI high‐cognition low‐Aβ progressed to AD (RP=0.00±0.01). Furthermore, we tested other cognitive assessments, such as CVLT, which provided similar or even statistically worse results in comparison to MMSE. The low‐cognition high‐Aβ group showed the fastest decline, with an annual rate of decline of 0.34 PACC scores, and an effect size of 0.75, compared to the high‐cognition, low‐Aβ group.

Conclusion
While cognitive performance alone is not sufficient for predicting MCI‐to‐AD progression, combining Aβ with MMSE cutoffs can enhance risk stratification, providing greater prognostic information and aiding in the design of clinical trials and therapeutic interventions for prodromal AD. This study also highlights the importance of using CL>44 to identify individuals at high risk of progressing to AD.

Background
Altered plasma sphingomyelin levels have been associated with Alzheimer's disease (AD), indicating disruptions in lipid metabolism and their potential link with AD pathogenesis. Meanwhile, tau‐related neuronal damage and neurodegeneration result in the release of neuronal tau protein into the bloodstream, reflected predominantly as plasma total‐Tau. Exploring the relationship between sphingomyelin metabolism and tau release before symptom onset could offer valuable insights into the biochemical pathways linked to AD and support the development of improved diagnostic and therapeutic strategies.

Method
The current study embraced plasma sample analysis using a Biocrates‐based targeted mass‐spectrometry platform to measure sphingomyelin levels and a single‐molecule array (Simoa) platform to quantify total‐Tau levels. The participants involved in this study were from the KARVIAH cohort —100 cognitively normal (CN; MoCA≥26 & MMSE≥26) older adults who underwent positron emission tomography (PET) to assess cortical amyloid‐beta (Aβ) load and were grouped as CN Aβ‐ (normal cortical‐Aβ load) and CN Aβ+ (higher cortical‐Aβ load). This study examined cross‐sectional correlations between plasma sphingomyelins and total‐Tau levels within each group. Sphingomyelins associated with total‐Tau were further analysed for their relationship with cortical‐Aβ load. All statistically significant correlations were assessed for correction for AD‐related confounding variables, containing gender, age, body mass index (BMI), and APOE ε4 status, with additional amendments for the false discovery rate (FDR).

Result
Statistically significant positive correlations between sphingomyelins and total‐Tau levels were observed exclusively in CN Aβ+ individuals. These associations remained robust after adjusting for AD‐related confounding variables and correcting for the FDR. Further analysis revealed significant inverse associations between total‐Tau‐associated sphingomyelins and cortical‐Aβ load, observed only in CN Aβ+ individuals, both with and without adjustment for confounding variables and FDR correction.

Conclusion
Elevated plasma total‐Tau levels were associated with higher plasma sphingomyelins exclusively in CN Aβ+ individuals, suggesting a link between tau release from neuronal damage and sphingomyelin‐associated biochemical pathways in the presymptomatic stage of AD. Furthermore, higher levels of total‐Tau‐associated sphingomyelins in plasma were correlated with lower cortical‐Aβ load in CN Aβ+ individuals, likely reflecting early sphingomyelin‐mediated compensatory mechanisms against AD pathogenesis. These findings highlight the potential of total‐Tau‐associated sphingomyelins as markers for understanding the mechanisms involved in presymptomatic AD.