Andrea Bozoki

Background and ObjectivesPlasma growth differentiation factor 15 (GDF15), a TGFbeta family cytokine, has been robustly identified from proteomics discovery work as linked to adverse brain aging; im...

Purpose: To compare longitudinal retinal and choroidal imaging findings in patients with mild cognitive impairment and cognitively normal control participants. Methods: Eyes of patients with mild cognitive impairment and cognitively normal controls were imaged using Zeiss Cirrus HD-5000 optical coherence tomography (OCT) with AngioPlex OCT angiography (OCTA). Macular vessel density and perfusion density were assessed. The Mini-Mental State Examination was administered at study entry and at least 18 months later. The velocity of change in OCT and OCTA parameters was compared between controls, nonamnestic patients with mild cognitive impairment, and amnestic patients with mild cognitive impairment. Results: Eighty-five eyes of 43 patients with mild cognitive impairment and 85 eyes of 43 control participants were analyzed at baseline, and 45% of those were analyzed at an average follow-up of 28 months. The entire cohort of patients with mild cognitive impairment demonstrated faster rates of decline in perfusion density (/year) in 3-mm circle ( P = .02) and ring ( P = .01) as well as in vessel density (mm ^-1 /year) in 3-mm circle ( P = .05) and ring ( P = .04) vs control participants. Amnestic patients with mild cognitive impairment demonstrated faster decline in perfusion density in 3-mm circle compared with nonamnestic patients with mild cognitive impairment ( P = .04) and control participants ( P = .007). Amnestic patients had a greater rate of decline in vessel density in 3-mm circle compared with nonamnestic patients ( P = .03) and control participants ( P = .02). The entire group of patients with mild cognitive impairment and the subgroup of nonamnestic patients had a faster rate of decline in perfusion density compared with control participants ( P = .01 and P < .001, respectively). Conclusions: The rate of decline in perfusion and vessel density was faster in the entire cohort of patients with mild cognitive impairment compared with control participants, with no significant change in Mini-Mental State Examination score. Amnestic patients with mild cognitive impairment demonstrated faster rates of decline in perfusion and vessel density compared with nonamnestic patients with mild cognitive impairment.

α-Synuclein (αSyn) seed amplification assays (SAAs) accurately detect αSyn pathology in dementia with Lewy bodies (DLB). Alzheimer disease (AD) copathology occurs in ∼70% of patients with DLB at autopsy. The association of different αSyn-SAA and AD biomarker profiles with clinical outcomes in DLB can aid prognostication and inform clinical trial design. This is a cross-sectional study of participants in the DLB Consortium (DLBC) study. The DLBC is a multicentered prospective longitudinal cohort study with participants recruited at tertiary care centers in the United States. CSF samples obtained within 1 year of baseline assessments were analyzed using αSyn-SAA and for β-amyloid (Aβ) 42/40, phosphorylated tau (p-tau) 181, and total tau (t-tau). Comparisons of baseline prevalence and degree of core DLB features, cognitive impairment, and hyposmia in participants with different CSF profiles were performed (AD-: normal Aβ42/40, p-tau181, and t-tau. AD+: Aβ42/40 abnormal, ±p-tau181, ±t-tau. Other non-AD: normal Aβ42/40 but elevated p-tau181 or t-tau levels). Longitudinal cognitive and motor declines were assessed in participants with different CSF profiles using linear mixed-effects models. Ninety-nine participants had CSF αSyn-SAA and AD biomarkers available for analysis (mean age 69.8 ± 6.6, 15% female); 68% (67/99) were αSyn-SAA+. αSyn-SAA+ participants had lower baseline Montreal Cognitive Assessment (MoCA) scores (19.0 ± 5.8 vs 21.6 ± 4.7, p = 0.03) and worse motor parkinsonism (Movement Disorder Society Unified Parkinson's Disease Rating Scale [MDS-UPDRS] part-III 32.3 ± 15.8 vs 22.1 ± 14.4, p = 0.003) and were more likely to be hyposmic (87% [46/53] vs 37% [10/27], p < 0.0001). αSyn-SAA+ participants had more progressive parkinsonism than αSyn-SAA- participants (MDS-UPDRS part-III: 3.70 points per year [95% CI 2.60-4.80] vs 0.47 points per year [95% CI -0.25 to 1.19], p < 0.0001). Of αSyn-SAA+ participants, 39% (26/67) were AD- and 52% (35/67) were AD+. αSyn-SAA+AD+ participants had greater declines in MoCA scores than αSyn-SAA+AD- participants (-3.24 points per year [95% CI -4.12 to -2.36] vs -1.19 points per year [95% CI -1.72 to -0.65], p < 0.0001). MDS-UPDRS part-I and part-II subscores increased more in αSyn-SAA+AD+ participants than in αSyn-SAA+AD- participants (2.72 points per year [95% CI 1.70-3.73] vs 1.21 points per year [95% CI 0.48-1.94], p = 0.03 and 4.60 points per year [95% CI 3.40-5.80] vs 1.93 points per year [95% CI 1.19-2.66], p < 0.001, respectively). In this well-characterized cohort of participants with clinically diagnosed DLB, αSyn-SAA positivity is associated with greater motor decline. AD copathology is common and confers additional risk of cognitive and functional decline.

Multimodal neuroimages, such as diffusion tensor imaging (DTI) and resting-state functional MRI (fMRI), offer complementary perspectives on brain activities by capturing structural or functional interactions among brain regions. While existing studies suggest that fusing these multimodal data helps detect abnormal brain activity caused by neurocognitive decline, they are generally implemented in Euclidean space and can’t effectively capture the intrinsic hierarchical organization of structural/functional brain networks. This paper presents a hyperbolic kernel graph fusion (HKGF) framework for neurocognitive decline analysis with multimodal neuroimages. It consists of a multimodal graph construction module, a graph representation learning module that encodes brain graphs in hyperbolic space through a family of hyperbolic kernel graph neural networks (HKGNNs), a cross-modality coupling module that enables effective multimodal data fusion, and a hyperbolic neural network for downstream predictions. Notably, HKGNNs represent graphs in hyperbolic space to capture both local and global dependencies among brain regions while preserving the hierarchical structure of brain networks. Extensive experiments involving over 4,000 subjects with DTI and/or fMRI data demonstrate the superiority of HKGF over state-of-the-art methods in two neurocognitive decline prediction tasks. The proposed HKGF is a general framework for multimodal data analysis, facilitating objective quantification of brain structural or functional connectivity changes associated with neurocognitive decline.

TMEM106B has been proposed as a modifier of disease risk in FTLD-TDP, particularly in GRN pathogenic variant carriers. Furthermore, TMEM106B has been investigated as a disease modifier in the context of healthy aging and across multiple neurodegenerative diseases. The objective of this study was to evaluate and compare the effect of TMEM106B on gray matter volume and cognition in each of the common genetic FTD groups and in patients with sporadic FTD. Participants were enrolled through the ARTFL/LEFFTDS Longitudinal Frontotemporal Lobar Degeneration (ALLFTD) study, which includes symptomatic and presymptomatic individuals with a pathogenic variant in C9orf72, GRN, MAPT, VCP, TBK1, TARDBP, symptomatic nonpathogenic variant carriers, and noncarrier family controls. All participants were genotyped for the TMEM106B rs1990622 SNP. Cross-sectionally, linear mixed-effects models were fitted to assess an association between TMEM106B and genetic group interaction with each outcome measure (gray matter volume and UDS3-EF for cognition), adjusting for education, age, sex, and CDR+NACC-FTLD sum of boxes. Subsequently, associations between TMEM106B and each outcome measure were investigated within the genetic group. For longitudinal modeling, linear mixed-effects models with time by TMEM106B predictor interactions were fitted. The minor allele of TMEM106B rs1990622, linked to a decreased risk of FTD, associated with greater gray matter volume in GRN pathogenic variant carriers under the recessive dosage model (N = 82, beta = 3.25, 95% CI [0.37-6.19], p = 0.034). This was most pronounced in the thalamus in the left hemisphere (beta = 0.03, 95% CI [0.01-0.06], p = 0.006), with a retained association when considering presymptomatic GRN pathogenic variant carriers only (N = 42, beta = 0.03, 95% CI [0.01-0.05], p = 0.003). The minor allele of TMEM106B rs1990622 also associated with greater cognitive scores among all C9orf72 pathogenic variant carriers (N = 229, beta = 0.36, 95% CI [0.05-0.066], p = 0.021) and in presymptomatic C9orf72 pathogenic variant carriers (N = 106, beta = 0.33, 95% CI [0.03-0.63], p = 0.036), under the recessive dosage model. We identified associations of TMEM106B with gray matter volume and cognition in the presence of GRN and C9orf72 pathogenic variants. The association of TMEM106B with outcomes of interest in presymptomatic GRN and C9orf72 pathogenic variant carriers could additionally reflect TMEM106B's effect on divergent pathophysiologic changes before the appearance of clinical symptoms.

Background: Diverse racial/ethnic representation in scientific studies is crucial for generalizability of results and development of potentially effective treatments. Racially diverse registries with research ready individuals have the potential to fill this gap. However, despite having signed up as part of a research registry and being invited to participate in clinical trials, many do not participate. To better understand research participation, the NC Registry for Brain Health surveyed its members to identify the characteristics of those likely to participate in research studies, their views on research, expectations when joining the group, and impediments to participation.Method: The NC Registry for Brain Health has been state‐funded since 2018 with goals of increasing access for all to information and research opportunities in brain health. Focused recruitment to under‐served and under‐represented groups across the State occurs through community outreach events sponsored by a five member university network. An email survey was sent to the 5,946 Registry members enrolled as of October 15, 2021.Result: Approximately 10% (n = 576) responded to this anonymized survey. Most were older women with high levels of formal education (Table 1). Despite the Registry’s demographics of 28% African American enrollees, only 8% of the survey respondents self‐identified as African Americans. Most individuals joined the Registry for the joint purpose of getting involved in research and receiving information about brain health (75%). When asked to describe what “brain health research” means, most reported studies involving dementia risk factors and memory testing. All participants had been approached to get involved in studies, but only 60% recalled getting an email invitation. The biggest impediment noted for participating in studies was inconvenient appointment times. The participants also wanted more available studies and more information about the opportunities.Conclusion: This study suggests that most participants in the Registry are seeking to be informed about brain health and have access to research opportunities at times convenient with their schedules. There was an observed gap between African Americans enrolled in the research registry and those who participated in the survey. Addressing gaps in research involvement of under‐represented groups in research is necessary to achieve inclusive representative science.

Alzheimer’s disease (AD) is a serious neurodegenerative disorder without a clear understanding of the etiology and pathophysiology. Recent experimental data has suggested neuronal excitation-inhibition (E-I) imbalance as an essential element and critical regulator of AD pathology, but E-I imbalance has not been systematically mapped out for either local or large-scale neuronal circuits in AD. By applying a Multiscale Neural Model Inversion (MNMI) framework to the resting-state functional MRI (rs-fMRI) data from the Alzheimer’s Disease Neuroimaging Initiative (ADNI), we identified brain regions with disrupted E-I balance based on impaired mesoscale excitatory and inhibitory connection strengths in a large network during AD progression. We observed that both intra-regional and inter-regional E-I balance is progressively disrupted from cognitively normal individuals, to mild cognitive impairment (MCI) and to AD, and E-I difference (or ratio) can be abnormally increased or decreased, depending on specific region. Also, we found that (local) inhibitory connections are more significantly impaired than excitatory ones and the strengths of the majority of connections are reduced in MCI and AD, leading to gradual decoupling of neural populations. Moreover, we revealed a core AD network comprised mainly of limbic and cingulate regions including the hippocampus, pallidum, putamen, nucleus accumbens, inferior temporal cortex and caudal anterior cingulate cortex (cACC). These brain regions exhibit consistent and stable E-I alterations across MCI and AD, and thus may represent early AD biomarkers and important therapeutic targets. Lastly, the E-I difference (or ratio) of multiple brain regions (precuneus, posterior cingulate cortex, pallium, cACC, putamen and hippocampus) was found to be significantly correlated with the Mini-Mental State Examination (MMSE) score, indicating that the degree of E-I impairment is behaviorally related to MCI/AD cognitive performance. Overall, our study constitutes the first attempt to delineate E-I imbalance in large-scale neuronal circuits during AD progression, which may facilitate the development of new treatment paradigms to restore pathological E-I balance in AD.