Normal Brain Aging Research Articles

Identification of Key Biomarkers and Pathways for Maintaining Cognitively Normal Brain Aging Based on Integrated Bioinformatics Analysis.

BackgroundGiven the arrival of the aging population has caused a series of social and economic problems, we aimed to explore the key genes underlying cognitively normal brain aging and its potential molecular mechanisms.MethodsGSE11882 was downloaded from Gene Expression Omnibus (GEO). The data from different brain regions were divided into aged and young groups for analysis. Co-expressed differentially expressed genes (DEGs) were screened. Functional analysis, protein–protein interaction (PPI) network, microRNA (miRNA)-gene, and transcription factor (TF)-gene networks were performed to identify hub genes and related molecular mechanisms. AlzData database was used to elucidate the expression of DEGs and hub genes in the aging brain. Animal studies were conducted to validate the hub genes.ResultsCo-expressed DEGs contained 7 upregulated and 87 downregulated genes. The enrichment analysis indicated DEGs were mainly involved in biological processes and pathways related to immune-inflammatory responses. From the PPI network, 10 hub genes were identified: C1QC, C1QA, C1QB, CD163, FCER1G, VSIG4, CD93, CD14, VWF, and CD44. CD44 and CD93 were the most targeted DEGs in the miRNA-gene network, and TIMP1, HLA-DRA, VWF, and FGF2 were the top four targeted DEGs in the TF-gene network. In AlzData database, the levels of CD44, CD93, and CD163 in patients with Alzheimer’s disease (AD) were significantly increased than those in normal controls. Meanwhile, in the brain tissues of cognitively normal mice, the expression of CD44, CD93, and CD163 in the aged group was significantly lower than those in the young group.ConclusionThe underlying molecular mechanisms for maintaining healthy brain aging are related to the decline of immune-inflammatory responses. CD44, CD93, and CD 163 are considered as potential biomarkers. This study provides more molecular evidence for maintaining cognitively normal brain aging.

Accumulation of systematic TPM1 mediates inflammation and neuronal remodeling by phosphorylating PKA and regulating the FABP5/NF-κB signaling pathway in the retina of aged mice.

The molecular mechanisms underlying functional decline during normal brain aging are poorly understood. Here, we identified the actin‐associated protein tropomyosin 1 (TPM1) as a new systemic pro‐aging factor associated with function deficits in normal aging retinas. Heterochronic parabiosis and blood plasma treatment confirmed that systemic factors regulated age‐related inflammatory responses and the ectopic dendritic sprouting of rod bipolar (RBC) and horizontal (HC) cells in the aging retina. Proteomic analysis revealed that TPM1 was a potential systemic molecule underlying structural and functional deficits in the aging retina. Recombinant TPM1 protein administration accelerated the activation of glial cells, the dendritic sprouting of RBCs and HCs and functional decline in the retina of young mice, whereas anti‐TPM1 neutralizing antibody treatment ameliorated age‐related structural and function changes in the retina of aged mice. Old mouse plasma (OMP) induced glial cell activation and the dendritic outgrowth of RBCs and HCs in young mice, and yet TMP1‐depleted OMP failed to reproduce the similar effect in young mice. These results confirmed that TPM1 was a systemic pro‐aging factor. Moreover, we demonstrated that systematic TPM1 was an immune‐related molecule, which elicited endogenous TPM1 expression and inflammation by phosphorylating PKA and regulating FABP5/NF‐κB signaling pathway in normal aging retinas. Interestingly, we observed TPM1 upregulation and the ectopic dendritic sprouting of RBCs and HCs in young mouse models of Alzheimer's disease, indicating a potential role of TPM1 in age‐related neurodegenerative diseases. Our data indicate that TPM1 could be targeted for combating the aging process.

Selective brain regional changes in lipid profile with human aging

Fatty acids are key components in the structural diversity of lipids and play a strategic role in the functional properties of lipids which determine the integrity of neuronal and glial cell membranes, the generation of lipid signaling mediators, and the chemical reactivity of acyl chains. The present study analyzes using gas chromatography the fatty acid profiles of 13 regions of the human central nervous system in healthy individuals ranging from 40 to 80 years old. The outcomes suggest the existence of general traits in fatty acid composition such as an average chain length of 18 carbon atoms, high monounsaturated fatty acid content, and predominance in polyunsaturated fatty acids of those of series n-6 over series n-3 which are shared by all brain regions regardless of age. Our results also show a general sustained and relatively well-preserved lipid profile throughout the adult lifespan in most studied regions (olive, upper vermis, substantia nigra, thalamus, hippocampus, putamen, caudate, occipital cortex, parietal cortex, entorhinal cortex, and frontal cortex) with minor changes that are region-dependent. In contrast, of particular relevance is the involvement of the inferior temporal cortex and cingulate cortex. It is proposed that during normal human brain aging, the lipid profile is resistant to changes with age in most human brain regions to ensure cell survival and function, but some particular regions involved in specific memory domains are greatly affected.

Gender-Related Differences in Regional Cerebral Glucose Metabolism in Normal Aging Brain.

Objectives: This study was aimed to investigate the gender-related differences of regional cerebral glucose metabolism in healthy people along the age using 18F-FDG PET/CT.Methods: We recruited 344 healthy volunteers, including 217 males and 127 females (age range: 40–89 years old). All subjects underwent fluorine-18 fluorodeoxyglucose(18F-FDG) positron emission tomography (PET). All the data were divided into four groups for every 10 years old. Each participant was carefully screened from PET, MR, and other examinations in order to exclude the abnormalities, such as neurodegenerative or psychiatric disorders, alcohol/abuse, cerebral vascular disorders, metabolic diseases like diabetes mellitus and hyperthyroidism, and other systemic malignancies. The 40–50 years old group was set as the baseline group. Statistical parametric mapping (SPM) analysis was employed to illustrate the differences among groups.Results: Compared to the baseline group, whether in a cohort or different gender groups, the decrease of brain glucose metabolism was shown in the bilateral frontal lobe, anterior cingulate gyrus, and the bilateral temporal lobe. In males, the regions of decreased metabolism were bilateral frontal lobe, caudate nucleus, and cingulate gyrus, whereas that of females were left occipital lobe, cerebellum, and the thalamus. However, the overall decrease of brain metabolism in men and women began from the age of 60s, an aggravated decrease from 70s was only observed in males.Conclusion: (1) An obviously decreased brain metabolism was found from 60 years old, especially in the bilateral frontal lobe, bilateral temporal lobe, and inferior cingulate gyrus; (2) We found specific brain metabolic differences between genders, including the caudate nucleus region in males and the occipital lobe region in females; and (3) The aging trend is different between genders.

Brain Morphometry and Cognitive Performance in Normal Brain Aging: Age- and Sex-Related Structural and Functional Changes.

BackgroundThe human brain structure undergoes considerable changes throughout life. Cognitive function can be affected either negatively or positively. It is challenging to segregate normal brain aging from the accelerated one.ObjectiveTo work out a descriptive model of brain structural and functional changes in normal aging.Materials and MethodsBy using voxel-based morphometry and lesion segmentation along with linear statistics and machine learning (ML), we analyzed the structural changes in the major brain compartments and modeled the dynamics of neurofunctional performance throughout life. We studied sex differences in lifelong dynamics of brain volumetric data with Mann-Whitney U-test. We tested the hypothesis that performance in some cognitive domains might decline as a linear function of age while other domains might have a non-linear dependence on it. We compared the volumetric changes in the major brain compartments with the dynamics of psychophysiological performance in 4 age groups. Then, we tested linear models of structural and functional decline for significant differences between the slopes in age groups with the T-test.ResultsWhite matter hyperintensities (WMH) are not the major structural determinant of the brain normal aging. They should be viewed as signs of a disease. There is a sex difference in the speed and/or in the onset of the gray matter atrophy. It either starts earlier or goes faster in males. Marked sex difference in the proportion of total cerebrospinal fluid (CSF) and intraventricular CSF (iCSF) justifies that elderly men are more prone to age-related brain atrophy than women of the same age.ConclusionThe article gives an overview and description of the conceptual structural changes in the brain compartments. The obtained data justify distinct patterns of age-related changes in the cognitive functions. Cross-life slowing of decision-making may follow the linear tendency of enlargement of the interhemispheric fissure because the center of task switching and inhibitory control is allocated within the medial wall of the frontal cortex, and its atrophy accounts for the expansion of the fissure. Free online tool at https://med-predict.com illustrates the tests and study results.

The Aged Striatum: Evidence of Molecular and Structural Changes Using a Longitudinal Multimodal Approach in Mice.

To study the aging human brain requires significant resources and time. Thus, mice models of aging can provide insight into changes in brain biological functions at a fraction of the time when compared to humans. This study aims to explore changes in dopamine D1 and D2 receptor availability and of gray matter density in striatum during aging in mice and to evaluate whether longitudinal imaging in mice may serve as a model for normal brain aging to complement cross-sectional research in humans. Mice underwent repeated structural magnetic resonance imaging (sMRI), and [11C]Raclopride and [11C]SCH23390 positron emission tomography (PET) was performed on a subset of aging mice. PET and sMRI data were analyzed by binding potential (BPND), voxel- and tensor-based morphometry (VBM and TBM, respectively). Longitudinal PET revealed a significant reduction in striatal BPND for D2 receptors over time, whereas no significant change was found for D1 receptors. sMRI indicated a significant increase in modulated gray matter density (mGMD) over time in striatum, with limited clusters showing decreased mGMD. Mouse [11C]Raclopride data is compatible with previous reports in human cross-sectional studies, suggesting that a natural loss of dopaminergic D2 receptors in striatum can be assessed in mice, reflecting estimates from humans. No changes in D1 were found, which may be attributed to altered [11C]SCH23390 kinetics in anesthetized mice, suggesting that this tracer is not yet able to replicate human findings. sMRI revealed a significant increase in mGMD. Although contrary to expectations, this increase in modulated GM density may be attributed to an age-related increase in non-neuronal cells.

Effects of Ketogenic Diet on Neuroinflammation in Neurodegenerative Diseases.

The ketogenic diet (KD) is a low-carbohydrate, high-fat and adequate-protein diet. As a diet mimicking fasting, it triggers the production of ketone bodies (KBs) and brings the body into a state of ketosis. Recent and accumulating studies on humans and animal models have shown that KD is beneficial to neurodegenerative diseases through modulating central and peripheral metabolism, mitochondrial function, inflammation, oxidative stress, autophagy, and the gut microbiome. Complicated interplay of metabolism, gut microbiome, and other mechanisms can regulate neuroinflammation in neurodegenerative diseases by activating multiple molecular and cellular pathways. In this review, we detail the physiological basis of the KD, its functions in regulating neuroinflammation, and its protective role in normal brain aging and neurodegenerative diseases, such as Alzheimer’s disease (AD), Parkinson’s disease (PD), amyotrophic lateral sclerosis (ALS), and Huntington’s disease (HD). We aimed to elucidate the underlying neuroinflammatory mechanisms of KD therapies in neurodegenerative diseases and provide novel insights into their application for neurodegenerative disease prevention and treatment.

Alzheimer's disease is an inherent, natural part of human brain aging: an integrated perspective.

Alzheimer disease is one of the most challenging demons in our society due to its very high prevalence and its clinical manifestations which cause deterioration of cognition, intelligence, and emotions - the very capacities that distinguish Homo sapiens from other animal species. Besides the personal, social, and economical costs, late stages of AD are vivid experiences for the family, relatives, friends, and general observers of the progressive ruin of an individual who turns into a being with lower mental and physical capacities than less evolved species. A human brain with healthy cognition, conscience, and emotions can succeed in dealing with most difficulties that life may pose. Without these capacities, the same person probably cannot. Due, in part, to this emotional impact, the absorbing study of AD has generated, over the years, a fascinating and complex story of theories, hypotheses, controversies, fashion swings, and passionate clashes, together with tremendous efforts and achievements geared to improve understanding of the pathogenesis and treatment of the disorder. Familal AD is rare and linked to altered genetic information associated with three genes. Sporadic AD (sAD) is much more common and multifactorial. A major point of clinical discussion has been, and still is, establishing the differences between brain aging and sAD. This is not a trivial question, as the neuropathological and molecular characteristics of normal brain aging and the first appearance of early stages of sAD-related pathology are not easily distinguishable in most individuals. Another important point is confidence in assigning responsibility for the beginning of sAD to a few triggering molecules, without considering the wide number of alterations that converge in the pathogenesis of aging and sAD. Genetic risk factors covering multiple molecular signals are increasing in number. In the same line, molecular pathways are altered at early stages of sAD pathology, currently grouped under the aegis of normal brain aging, only to increase massively at advanced stages of the process. Sporadic AD is here considered an inherent, natural part of human brain aging, which is prevalent in all humans, and variably present or not in a few individuals in other species. The progression of the process has devastating effects in a relatively low percentage of human beings eventually evolving to dementia. The continuum of brain aging and sAD implies the search for a different approach in the study of human brain aging at the first stages of the biological process, and advances in the use of new technologies aimed at slowing down the molecular defects underlying human brain aging and sAD at the outset, and transfering information and tasks to AI and coordinated devices.

Multiscale Brain Aging in the Context of Neurodegeneration and Alzheimer's Disease

The brain, with a diverse array of specialized cells, regional substructures, and a relatively isolated microenvironment, represents a uniquely challenging organ system for aging research. The brain can experience physical trauma, interact with the periphery, and is responsible for cognitive and behavioral modifications that can feed back into the molecular processes of aging both within and external to the brain. Advances to our understanding and ability to intervene in the complexity that personifies brain aging and associated neurodegeneration will require integrated, multiscale approaches operating in tandem. Therefore, we have organized this symposium to highlight promising new approaches to study brain aging through the lens of multiple biological levels of organization. We will provide insight not only into normal brain aging, but will also suggest key spurious processes that may drive neurodegeneration and functional decline.

Emotional Well-Being Human Studies

Early evidence indicates an association between EWB and underlying brain processes, and that those processes change with both normal and pathological brain aging. However, the nature of these associations, the mechanisms by which EWB and its component domains change with brain aging, and how those changes may be associated with common neuropathologies in ADRD, are largely unexplored. We propose an appraisal-adaptation model in understanding relationships between EWB and ADRD. For human models, we encourage the use of well-established measures that directly assess eudaimonic and hedonic EWB, including abnormal scenarios (e.g., neuropsychiatric symptoms, anhedonia, loneliness, etc.), as well as older adults with exceptional cognition (i.e., superagers or supernormals). Dr. Lin will review premises associated with the appraisal-adaptation model in conducting human research on EWB, aging, and ADRD.

New Brain Aging Center

Evidence indicates an association between emotional well-being (EWB) and underlying brain processes, and that those processes change with both normal and pathological brain aging. However, the nature of these associations, the mechanisms by which EWB and its component domains change with brain aging, and how those changes may be associated with common neuropathologies like Alzheimer’s disease and related dementias (ADRD), are largely unexplored. The NIA-funded Network for Emotional Well-being and Brain Aging (NEW Brain Aging) has the goal of developing a nationwide community of investigators dedicated to research that identifies and tests mechanisms by which brain aging influences EWB and how EWB may impact risk for and progression of ADRD. Synthesizing human and animal literature, our premise is that relationships between EWB and ADRD are bidirectional – normal and pathological changes in aging brain influence EWB and EWB contributes to brain health and illness, such as ADRD. NEW Brain Aging will identify and coalesce resources for interested investigators and provide pilot funding opportunities to stimulate research and development of the field. Component presentations of this symposium will include (1) an overview by Dr. Robert Kaplan of the current state of research on EWB; (2) the role of animal studies (Kuan Hong Wang) and (3) human subjects research (Feng Vankee Lin) in EWB and aging; and (4) design of NEW Brain Aging and resources it will provide (Yeates Conwell). Janine Simmons will explain NIA’s vision for EWB research and lead open discussion.

Transcriptomic Changes Highly Similar to Alzheimer's Disease Are Observed in a Subpopulation of Individuals During Normal Brain Aging.

Aging is a major risk factor for late-onset Alzheimer’s disease (LOAD). How aging contributes to the development of LOAD remains elusive. In this study, we examined multiple large-scale transcriptomic datasets from both normal aging and LOAD brains to understand the molecular interconnection between aging and LOAD. We found that shared gene expression changes between aging and LOAD are mostly seen in the hippocampal and several cortical regions. In the hippocampus, the expression of phosphoprotein, alternative splicing and cytoskeleton genes are commonly changed in both aging and AD, while synapse, ion transport, and synaptic vesicle genes are commonly down-regulated. Aging-specific changes are associated with acetylation and methylation, while LOAD-specific changes are more related to glycoprotein (both up- and down-regulations), inflammatory response (up-regulation), myelin sheath and lipoprotein (down-regulation). We also found that normal aging brain transcriptomes from relatively young donors (45–70 years old) clustered into several subgroups and some subgroups showed gene expression changes highly similar to those seen in LOAD brains. Using brain transcriptomic datasets from another cohort of older individuals (>70 years), we found that samples from cognitively normal older individuals clustered with the “healthy aging” subgroup while AD samples mainly clustered with the “AD similar” subgroups. This may imply that individuals in the healthy aging subgroup will likely remain cognitively normal when they become older and vice versa. In summary, our results suggest that on the transcriptome level, aging and LOAD have strong interconnections in some brain regions in a subpopulation of cognitively normal aging individuals. This supports the theory that the initiation of LOAD occurs decades earlier than the manifestation of clinical phenotype and it may be essential to closely study the “normal brain aging” to identify the very early molecular events that may lead to LOAD development.

Cortical remodeling across the lifespan in healthy brain reveals structural network vulnerability to neurodegeneration

AbstractBackgroundAging is the main risk factor for most of the neurodegenerative diseases. The aim of this study was to investigate typical cortical thinning changes across lifespan in the healthy brain revealing structural network vulnerability to neurodegeneration.MethodThe cohort included 128 healthy individuals aged 20‐85 years that underwent an MRI scan. Structural T1‐weighted images were used to estimate vertex‐wise cortical thickness maps, then grouped into 83 regions. For each region, cortical thickness trajectory with advancing age was estimated, including sex as covariate. Additionally, all regions were ranked based on their relative thickness at the end of the observed lifetime, assessing regional changes over time. Finally, regional mean thickness was correlated with relative change over time.ResultThe highest cortical thinning was observed in the temporal lobe (parahippocampal, entorhinal, superior and middle temporal and fusiform), in the frontal lobe (lateral orbitofrontal, superior and inferior frontal and rostral anterior cingulate), in the parietal lobe (the isthmus of cingulate, precuneus, supramarginal and inferior parietal) and in the insular cortex. Interestingly, occipital regions (cuneus, lateral occipital, lingual, pericalcarine), and motor and premotor areas (precentral, postcentral and paracentral regions) showed the least cortical thickness change relative to the whole brain. Finally, positive correlation was found between mean regional thickness and its relative change over time.ConclusionThis study highlights structural vulnerability of brain regions to aging. Furthermore, results provide information concerning trajectories of normal brain aging, identifying those areas that might be more vulnerable to the attack of neurodegeneration. Supported by: European Research Council (StG‐2016_714388_NeuroTRACK).

Longitudinal White Matter Damage Evolution in Parkinson's Disease.

White matter hyperintensities (WMHs) have a role in cognitive impairment in normal brain aging, while the effect on Parkinson's disease (PD) progression is still controversial. To investigate the longitudinal evolution of micro- and macrostructural damage of cerebral white matter (WM) and its relationship with the clinical picture in PD. A total of 154 PD patients underwent clinical, cognitive, and magnetic resonance imaging (MRI) assessment once a year for up to 4 years. Sixty healthy controls underwent the same protocol at baseline. WMHs were identified and total WMH volume was measured. WMHs were also used as exclusion masks to define normal-appearing white matter (NAWM). Using tract-based spatial statistics, diffusion tensor (DT) MRI metrics of whole-brain WM and NAWM were obtained. Linear mixed-effects models defined the longitudinal evolution and association between variables. WM alterations were tested as risk factors of disease progression using linear regression and Cox proportional hazards models. At baseline, PD patients showed alterations of all DT MRI measures compared to controls. Longitudinally, DT MRI measures did not vary significantly and no association with clinical variables was found. WMH volume changed over time and was associated with impairment in global cognition, executive functions, and language. Baseline WMH volume was a moderate risk factor for progression to mild cognitive impairment. Our study suggests an association between WMHs and cognitive deterioration in PD, whereas WM microstructural damage is a negligible contributor to clinical deterioration. WMHs assessed by MRI can provide an important tool for monitoring the development of cognitive impairment in PD patients. © 2021 International Parkinson and Movement Disorder Society.

Proportional Changes in Cognitive Subdomains During Normal Brain Aging

Background: Neuroscience lacks a reliable method of screening the early stages of dementia.Objective: To improve the diagnostics of age-related cognitive functions by developing insight into the proportionality of age-related changes in cognitive subdomains.Materials and Methods: We composed a battery of psychophysiological tests and collected an open-access psychophysiological outcomes of brain atrophy (POBA) dataset by testing individuals without dementia. To extend the utility of machine learning (ML) classification in cognitive studies, we proposed estimates of the disproportional changes in cognitive functions: an index of simple reaction time to decision-making time (ISD), ISD with the accuracy performance (ISDA), and an index of performance in simple and complex visual-motor reaction with account for accuracy (ISCA). Studying the distribution of the values of the indices over age allowed us to verify whether diverse cognitive functions decline equally throughout life or there is a divergence in age-related cognitive changes.Results: Unsupervised ML clustering shows that the optimal number of homogeneous age groups is four. The sample is segregated into the following age-groups: Adolescents ∈ [0, 20), Young adults ∈ [20, 40), Midlife adults ∈ [40, 60) and Older adults ≥60 year of age. For ISD, ISDA, and ISCA values, only the median of the Adolescents group is different from that of the other three age-groups sharing a similar distribution pattern (p > 0.01). After neurodevelopment and maturation, the indices preserve almost constant values with a slight trend toward functional decline. The reaction to a moving object (RMO) test results (RMO_mean) follow another tendency. The Midlife adults group's median significantly differs from the remaining three age subsamples (p < 0.01). No general trend in age-related changes of this dependent variable is observed. For all the data (ISD, ISDA, ISCA, and RMO_mean), Levene's test reveals no significant changes of the variances in age-groups (p > 0.05). Homoscedasticity also supports our assumption about a linear dependency between the observed features and age.Conclusion: In healthy brain aging, there are proportional age-related changes in the time estimates of information processing speed and inhibitory control in task switching. Future studies should test patients with dementia to determine whether the changes of the aforementioned indicators follow different patterns.

A fiber-deprived diet causes cognitive impairment and hippocampal microglia-mediated synaptic loss through the gut microbiota and metabolites

BackgroundCognitive impairment, an increasing mental health issue, is a core feature of the aging brain and neurodegenerative diseases. Industrialized nations especially, have experienced a marked decrease in dietary fiber intake, but the potential mechanism linking low fiber intake and cognitive impairment is poorly understood. Emerging research reported that the diversity of gut microbiota in Western populations is significantly reduced. However, it is unknown whether a fiber-deficient diet (which alters gut microbiota) could impair cognition and brain functional elements through the gut-brain axis.ResultsIn this study, a mouse model of long-term (15 weeks) dietary fiber deficiency (FD) was used to mimic a sustained low fiber intake in humans. We found that FD mice showed impaired cognition, including deficits in object location memory, temporal order memory, and the ability to perform daily living activities. The hippocampal synaptic ultrastructure was damaged in FD mice, characterized by widened synaptic clefts and thinned postsynaptic densities. A hippocampal proteomic analysis further identified a deficit of CaMKIId and its associated synaptic proteins (including GAP43 and SV2C) in the FD mice, along with neuroinflammation and microglial engulfment of synapses. The FD mice also exhibited gut microbiota dysbiosis (decreased Bacteroidetes and increased Proteobacteria), which was significantly associated with the cognitive deficits. Of note, a rapid differentiating microbiota change was observed in the mice with a short-term FD diet (7 days) before cognitive impairment, highlighting a possible causal impact of the gut microbiota profile on cognitive outcomes. Moreover, the FD diet compromised the intestinal barrier and reduced short-chain fatty acid (SCFA) production. We exploit these findings for SCFA receptor knockout mice and oral SCFA supplementation that verified SCFA playing a critical role linking the altered gut microbiota and cognitive impairment.ConclusionsThis study, for the first time, reports that a fiber-deprived diet leads to cognitive impairment through altering the gut microbiota-hippocampal axis, which is pathologically distinct from normal brain aging. These findings alert the adverse impact of dietary fiber deficiency on brain function, and highlight an increase in fiber intake as a nutritional strategy to reduce the risk of developing diet-associated cognitive decline and neurodegenerative diseases.2_wgC4CVAKtYwCNZbg5Y4xVideo

Neurofilament Proteins as Biomarkers to Monitor Neurological Diseases and the Efficacy of Therapies.

Biomarkers of neurodegeneration and neuronal injury have the potential to improve diagnostic accuracy, disease monitoring, prognosis, and measure treatment efficacy. Neurofilament proteins (NfPs) are well suited as biomarkers in these contexts because they are major neuron-specific components that maintain structural integrity and are sensitive to neurodegeneration and neuronal injury across a wide range of neurologic diseases. Low levels of NfPs are constantly released from neurons into the extracellular space and ultimately reach the cerebrospinal fluid (CSF) and blood under physiological conditions throughout normal brain development, maturation, and aging. NfP levels in CSF and blood rise above normal in response to neuronal injury and neurodegeneration independently of cause. NfPs in CSF measured by lumbar puncture are about 40-fold more concentrated than in blood in healthy individuals. New ultra-sensitive methods now allow minimally invasive measurement of these low levels of NfPs in serum or plasma to track disease onset and progression in neurological disorders or nervous system injury and assess responses to therapeutic interventions. Any of the five Nf subunits – neurofilament light chain (NfL), neurofilament medium chain (NfM), neurofilament heavy chain (NfH), alpha-internexin (INA) and peripherin (PRPH) may be altered in a given neuropathological condition. In familial and sporadic Alzheimer’s disease (AD), plasma NfL levels may rise as early as 22 years before clinical onset in familial AD and 10 years before sporadic AD. The major determinants of elevated levels of NfPs and degradation fragments in CSF and blood are the magnitude of damaged or degenerating axons of fiber tracks, the affected axon caliber sizes and the rate of release of NfP and fragments at different stages of a given neurological disease or condition directly or indirectly affecting central nervous system (CNS) and/or peripheral nervous system (PNS). NfPs are rapidly emerging as transformative blood biomarkers in neurology providing novel insights into a wide range of neurological diseases and advancing clinical trials. Here we summarize the current understanding of intracellular NfP physiology, pathophysiology and extracellular kinetics of NfPs in biofluids and review the value and limitations of NfPs and degradation fragments as biomarkers of neurodegeneration and neuronal injury.

Multivariate semi-blind deconvolution of fMRI time series

Whole brain estimation of the haemodynamic response function (HRF) in functional magnetic resonance imaging (fMRI) is critical to get insight on the global status of the neurovascular coupling of an individual in healthy or pathological condition. Most of existing approaches in the literature works on task-fMRI data and relies on the experimental paradigm as a surrogate of neural activity, hence remaining inoperative on resting-stage fMRI (rs-fMRI) data. To cope with this issue, recent works have performed either a two-step analysis to detect large neural events and then characterize the HRF shape or a joint estimation of both the neural and haemodynamic components in an univariate fashion. In this work, we express the neural activity signals as a combination of piece-wise constant temporal atoms associated with sparse spatial maps and introduce an haemodynamic parcellation of the brain featuring a temporally dilated version of a given HRF model in each parcel with unknown dilation parameters. We formulate the joint estimation of the HRF shapes and spatio-temporal neural representations as a multivariate semi-blind deconvolution problem in a paradigm-free setting and introduce constraints inspired from the dictionary learning literature to ease its identifiability. A fast alternating minimization algorithm, along with its efficient implementation, is proposed and validated on both synthetic and real rs-fMRI data at the subject level. To demonstrate its significance at the population level, we apply this new framework to the UK Biobank data set, first for the discrimination of haemodynamic territories between balanced groups (n=24 individuals in each) patients with an history of stroke and healthy controls and second, for the analysis of normal aging on the neurovascular coupling. Overall, we statistically demonstrate that a pathology like stroke or a condition like normal brain aging induce longer haemodynamic delays in certain brain areas (e.g. Willis polygon, occipital, temporal and frontal cortices) and that this haemodynamic feature may be predictive with an accuracy of 74 % of the individual’s age in a supervised classification task performed on n=459 subjects.

The progression of disorder-specific brain pattern expression in schizophrenia over 9 years

Age plays a crucial role in the performance of schizophrenia vs. controls (SZ-HC) neuroimaging-based machine learning (ML) models as the accuracy of identifying first-episode psychosis from controls is poor compared to chronic patients. Resolving whether this finding reflects longitudinal progression in a disorder-specific brain pattern or a systematic but non-disorder-specific deviation from a normal brain aging (BA) trajectory in schizophrenia would help the clinical translation of diagnostic ML models. We trained two ML models on structural MRI data: an SZ-HC model based on 70 schizophrenia patients and 74 controls and a BA model (based on 561 healthy individuals, age range = 66 years). We then investigated the two models’ predictions in the naturalistic longitudinal Northern Finland Birth Cohort 1966 (NFBC1966) following 29 schizophrenia and 61 controls for nine years. The SZ-HC model’s schizophrenia-specificity was further assessed by utilizing independent validation (62 schizophrenia, 95 controls) and depression samples (203 depression, 203 controls). We found better performance at the NFBC1966 follow-up (sensitivity = 75.9%, specificity = 83.6%) compared to the baseline (sensitivity = 58.6%, specificity = 86.9%). This finding resulted from progression in disorder-specific pattern expression in schizophrenia and was not explained by concomitant acceleration of brain aging. The disorder-specific pattern’s progression reflected longitudinal changes in cognition, outcomes, and local brain changes, while BA captured treatment-related and global brain alterations. The SZ-HC model was also generalizable to independent schizophrenia validation samples but classified depression as control subjects. Our research underlines the importance of taking account of longitudinal progression in a disorder-specific pattern in schizophrenia when developing ML classifiers for different age groups.

Classification-Biased Apparent Brain Age for the Prediction of Alzheimer's Disease.

Machine Learning methods are often adopted to infer useful biomarkers for the early diagnosis of many neurodegenerative diseases and, in general, of neuroanatomical ageing. Some of these methods estimate the subject age from morphological brain data, which is then indicated as “brain age”. The difference between such a predicted brain age and the actual chronological age of a subject can be used as an indication of a pathological deviation from normal brain ageing. An important use of the brain age model as biomarker is the prediction of Alzheimer's disease (AD) from structural Magnetic Resonance Imaging (MRI). Many different machine learning approaches have been applied to this specific predictive task, some of which have achieved high accuracy at the expense of the descriptiveness of the model. This work investigates an appropriate combination of data science techniques and linear models to provide, at the same time, high accuracy and good descriptiveness. The proposed method is based on a data workflow that include typical data science methods, such as outliers detection, feature selection, linear regression, and logistic regression. In particular, a novel inductive bias is introduced in the regression model, which is aimed at improving the accuracy and the specificity of the classification task. The method is compared to other machine learning approaches for AD classification based on morphological brain data with and without the use of the brain age, including Support Vector Machines and Deep Neural Networks. This study adopts brain MRI scans of 1, 901 subjects which have been acquired from three repositories (ADNI, AIBL, and IXI). A predictive model based only on the proposed apparent brain age and the chronological age has an accuracy of 88% and 92%, respectively, for male and female subjects, in a repeated cross-validation analysis, thus achieving a comparable or superior performance than state of the art machine learning methods. The advantage of the proposed method is that it maintains the morphological semantics of the input space throughout the regression and classification tasks. The accurate predictive model is also highly descriptive and can be used to generate potentially useful insights on the predictions.