| Perception of social experiences and cortical thickness change together throughout early adolescence: findings from the ABCD cohort | 10.15154/1524763 | Early adolescence is a dynamic period of social and cortical development amidst rapid hormonal and puberty changes. We examined how differences and changes in positive social experiences and cortical thickness co-develop from age 9-11 and 11-13 years in the ABCD cohort (N~12,000). We used Bivariate Latent Change Score Models to capture cortical development (modelling mean whole-brain cortical thickness) and positive social experiences (modelling caregiver monitoring, family cohesion, prosocial behaviour, number of friends, school engagement, school involvement, and neighborhood safety). We found that positive social experiences decreased between age 9-11 years (baseline) and 11-13 years (2-year-follow-up), indicating that social experiences were perceived as less positive over time. We found evidence for correlated change, such that a greater reduction in positive social experiences was associated with a greater reduction in cortical thickness (est=5.23, SE=1.31, p<.001, standardized coefficient=.08), which did not differ between males and females in early and late puberty stages. We found mixed evidence for sex-specific relationships between puberty stage and social experiences. The evidence supports a transactional model of development, in that positive social experiences and cortical thickness change together throughout early adolescence. The findings also highlight the importance of supporting youth in early adolescence through school transitions. | 725/13803 | Secondary Analysis | Shared |
| Lifespan development of brain asymmetry | 10.15154/3er3-dc69 | Lateralization is a fundamental principle of structural brain organization. In vivo imaging of brain asymmetry is essential for deciphering lateralized brain functions and their disruption in neurodevelopmental and neurodegenerative disorders. Here, we present a normative framework for benchmarking brain asymmetry across the lifespan, developed from an aggregated sample of 128 primary neuroimaging studies, including 177,701 scans from 138,231 individuals, jointly spanning the age range from 20 post menstrual weeks to 102 years. This resource includes comprehensive, hemisphere-specific brain growth charts for multiple neuroimaging phenotypes: regional cortical grey matter volume, thickness, surface area, and subcortical volumes. Our findings reveal distinct spatial patterns of asymmetry, with early leftward asymmetry observed in association cortices and late rightward asymmetry in sensory regions. These trajectories support theories of the neuroplasticity of asymmetry and the role of both genetic and environmental factors in shaping brain lateralization. Additionally, we provide tools to generate asymmetry centile scores, which allow the quantification of individual deviations from typical asymmetry throughout the lifespan and can be applied to unseen data or clinical populations. We demonstrate the utility of these models by highlighting group-level differences in asymmetry in autism spectrum disorder, schizophrenia, and Alzheimer’s disease, and exploring genetic correlations with hemispheric specialization. To facilitate further research, we have made this normative framework freely available as an interactive open-access resource (upon publication), offering an essential tool to advance both basic and clinical neuroscience. | 725/8552 | Secondary Analysis | Shared |
| ComBatLS: A location- and scale-preserving method for multi-site image harmonization | 10.15154/sr0j-g796 | Recent work has leveraged massive datasets and advanced harmonization methods to construct normative models of neuroanatomical features and benchmark individuals’ morphology. However, current harmonization tools do not preserve the effects of biological covariates including sex and age on features’ variances; this failure may induce error in normative scores, particularly when such factors are distributed unequally across sites. Here, we introduce a new extension of the popular ComBat harmonization method, ComBatLS, that preserves biological variance in features’ locations and scales. We use UK Biobank data to show that ComBatLS robustly replicates individuals’ normative scores better than other ComBat methods when subjects are assigned to sex-imbalanced synthetic “sites”. Additionally, we demonstrate that ComBatLS significantly reduces sex biases in normative scores compared to traditional methods. Finally, we show that ComBatLS successfully harmonizes consortium data collected across over 50 studies. R implementation of ComBatLS is available at https://github.com/andy1764/ComBatFamily. | 725/6514 | Secondary Analysis | Shared |
| Verbal Learning Harmonization | 10.15154/zb4d-s646 | Motivation: Auditory verbal learning tasks (AVLTs) are a core component of neuropsychological assessment, but the variety of AVLTs in common use makes it difficult to compare scores across instruments. This limits integration of research findings. The objective of this study was to derive and disseminate crosswalks that directly equate raw scores across common AVLTs.
Methods: A large, international repository of raw AVLT data was compiled, and a multisite mega study analysis was conducted. Empirical Bayes harmonization was used to isolate and remove site effects, followed by linear models which adjusted for covariates, including age, sex, education, and race/ethnicity. After corrections, a continuous item response theory (IRT) model was then used to estimate each individual subject’s latent verbal learning ability while accounting for different item difficulties.
Results: We aggregated raw data from studies of clinical samples and healthy controls from around the world that measured at least one verbal learning task. After applying exclusion criteria, the final sample was comprised of N = 10,505 individuals with and without history of traumatic brain injury from 53 studies above the age of 16 years who were tested on the California Verbal Learning Test (CVLT), Rey Auditory Verbal Learning Test (RAVLT), or the Hopkins Verbal Learning Test-Revised (HVLT). Harmonization significantly reduced inter-site variance by 37% while preserving covariate effects for further study. The effects of age, sex, and education on scores were reported and were found to be consistent across all AVLTs. Crosswalks were created by linking scores of individuals with the same verbal learning ability across AVLTs. The derived conversions agreed with held-out data of dually-administered tests.
Conclusion: This study reports the co-calibration and validation of methods to harmonize raw scores across three common verbal learning instruments. Moreover, we developed a free online tool for cross-assessment raw score conversion. These methods address longstanding data compatibility issues for AVLTs, and offer perspectives on how large-scale data harmonization initiatives can increase the robustness and reproducibility of research and findings across the behavioral sciences.
| 720/3745 | Secondary Analysis | Shared |
| Evidence for embracing normative modeling | 10.15154/9c5r-0h50 | In this work, we expand the normative model repository introduced in Rutherford et al., 2022a to include normative models charting lifespan trajectories of structural surface area and brain functional connectivity, measured using two unique resting-state network atlases (Yeo-17 and Smith-10), and an updated online platform for transferring these models to new data sources. We showcase the value of these models with a head-to-head comparison between the features output by normative modeling and raw data features in several benchmarking tasks: mass univariate group difference testing (schizophrenia versus control), classification (schizophrenia versus control), and regression (predicting general cognitive ability). Across all benchmarks, we show the advantage of using normative modeling features, with the strongest statistically significant results demonstrated in the group difference testing and classification tasks. We intend for these accessible resources to facilitate the wider adoption of normative modeling across the neuroimaging community. | 725/2599 | Secondary Analysis | Shared |
| Temporal and Spatial Scales of Human Resting-state Cortical Activity Across the Lifespan | 10.15154/5f0d-z111 | Sensorimotor and cognitive abilities undergo substantial changes throughout the human lifespan, but the corresponding changes in the functional properties of cortical network remain poorly understood. This can be studied using temporal and spatial scales of functional magnetic resonance imaging (fMRI) signals, which provide a robust description of the topological structure and temporal dynamics of neural activity. For example, timescales of resting-state fMRI signals can parsimoniously predict a significant amount of the individual variability in functional connectivity networks identified in adult human brains. In the present study, we quantified and compared temporal and spatial scales in resting-state fMRI data collected from 2,352 subjects between the ages of 5 and 100 in Developmental, Young Adult, and Aging datasets from Human Connectome Project. For most cortical regions, we found that both temporal and spatial scales largely decreased with age across most cortical areas throughout the lifespan, with the visual cortex and the limbic network consistently showing the largest and smallest scales, respectively. For some prefrontal regions, however, these two scales displayed non-monotonic trajectories during adolescence and peaked around the same time during adolescence and decreasing throughout the rest of the lifespan. We also found that cortical myelination increased monotonically throughout the lifespan, and its rate of change was significantly correlated with the changes in both temporal and spatial scales across different cortical regions in adulthood. These findings suggest that temporal and spatial scales in fMRI signals, as well as cortical myelination, are closely coordinated during both development and aging. | 725/1377 | Secondary Analysis | Shared |
| The structure of neuroanatomical variation within bilinguals | 10.15154/1528104 | We have developed this CVAE method that was useful for Autism (Aglinskas et al 2022). We think it might be useful for bilingualism research because bilingualism is similarly variable. We need access to this database to test our ideas. | 723/1375 | Secondary Analysis | Shared |
| Functional Brain Network Dynamics Mediate the Relationship between Female Reproductive Aging and Perceived Hostility: Relevance to Neuropsychiatric Disorders | 10.15154/g3hb-nf28 | Premature reproductive aging is linked to heightened stress sensitivity and psychological maladjustment across the life course. However, the brain dynamics underlying this relationship are poorly understood. To address this issue, we analysed multimodal data from female participants in the Adolescent Brain and Cognitive Development (longitudinal, N=441; aged 9-12 years) and Human Connectome-Aging (cross-sectional, N =130; aged 36-60 years) studies. Age-specific intrinsic functional brain network dynamics mediated the link between reproductive aging and perceptions of greater interpersonal adversity. The adolescent profile overlapped areas of greater glutamatergic and dopaminergic receptor density, while the middle-aged profile was concentrated in visual, attentional and default mode networks. The two profiles showed opposite relationships with patterns of functional neural network variability and cortical atrophy observed in psychosis vs major depressive disorder. Our findings underscore the divergent patterns of brain aging linked to reproductive maturation vs senescence, which may explain developmentally specific vulnerability to distinct disorders. | 130/849 | Secondary Analysis | Shared |
| Age-related differences in network controllability are mitigated by redundancy in large-scale brain networks | 10.15154/j38f-bg57 | The aging brain undergoes major changes in its topology. The mechanisms by which the brain mitigates age-associated changes in topology to maintain robust control of brain networks are unknown. Here we use diffusion MRI data from cognitively intact participants (n = 480, ages 40-90) to study age-associated differences in the average controllability of structural brain networks, topological features that could mitigate these differences, and the overall effect on cognitive function. We find age-associated declines in average controllability in control hubs and large-scale networks, particularly within the frontoparietal control and default mode networks. Further, we find that redundancy, a hypothesized mechanism of reserve, quantified via the assessment of multi-step paths within networks, mitigates the effects of topological differences on average network controllability. Lastly, we discover that average network controllability, redundancy, and grey matter volume, each uniquely contribute to predictive models of cognitive function. In sum, our results highlight the importance of redundancy for robust control of brain networks and in cognitive function in healthy-aging.
| 725/725 | Secondary Analysis | Shared |
| A cross-cohort replicable and heritable latent dimension linking behaviour to multi-featured brain structure | 10.15154/1527877 | Identifying the complex association between inter-individual variability in brain structure and in behaviour is challenging, requiring large cohorts, multivariate methods, out-of-sample validation and, ideally, out-of-cohort replication. Moreover, the influence of nature vs nurture on such associations could be better understood through heritability analysis. Here, we analysed associations between brain structure and behaviour using regularized canonical correlation analysis and a recently proposed machine learning framework that tests the generalisability of such associations. We linked behaviour (spanning cognition, emotion, and alertness) and multi-featured brain structure (grey matter volume, cortical thickness, and surface area). The replicability of such brain-behaviour associations was assessed in two large and independent cohorts. The load of genetic factors on these latent dimensions was analysed with heritability and genetic correlation. We found one heritable and replicable latent dimension. This latent dimension was positively associated with cognitive-control/executive-functions and positive affect, and negatively associated with impulsivity and negative affect. This behavioural profile was related to brain structural variability in areas typically associated with higher cognitive functions, as well as with areas typically associated with sensorimotor functions. These results revealed a major axis of interindividual variability in behaviour that links to a whole-brain structural pattern. | 725/725 | Primary Analysis | Shared |
| Cortical myelin profile variations in healthy aging brain: A T1w/T2w ratio study | 10.15154/1528709 | Demyelination is observed in both healthy aging and age-related neurodegenerative disorders. While the significance of myelin within the cortex is well acknowledged, studies focused on intracortical demyelination and depth-specific structural alterations in normal aging are lacking. Using the recently available Human Connectome Project Aging dataset, we investigated intracortical myelin in a normal aging population using the T1w/T2w ratio. To capture the fine changes across cortical depths, we employed a surface-based approach by constructing cortical profiles traveling perpendicularly through the cortical ribbon and sampling T1w/T2w values. The curvatures of T1w/T2w cortical profiles may be influenced by differences in local myeloarchitecture and other tissue properties, which are known to vary across cortical regions. To quantify the shape of these profiles, we parametrized the level of curvature using a nonlinearity index (NLI) that measures the deviation of the profile from a straight line. We showed that NLI exhibited a steep decline in aging that was independent of local cortical thinning. Further examination of the profiles revealed that lower T1w/T2w near the gray-white matter boundary and superficial cortical depths were major contributors to the apparent NLI variations with age. These findings suggest that demyelination and changes in other T1w/T2w related tissue properties in normal aging may be depth-specific and highlight the potential of NLI as a unique marker of microstructural alterations within the cerebral cortex. | 725/725 | Secondary Analysis | Shared |
| Cortical-structural insights into empathy and its correlates | 10.15154/1asw-j733 | Insights into the brain structure underlying empathy – understanding others’ mental states (cognitive empathy) and responding to them with an appropriate emotion (affective empathy) – are fragmented. Here, we integrate cortical-structural and purely behavioural insights into empathy and its correlates across two MRI studies of cortical thickness (CT) and surface area (SA), critical and complementary constituents of the cerebral cortex. Chapter 1 reviews the relevant literature. Chapter 2 presents a study on CT and SA in relation to empathy and empathising-systemising, and the effects of autism, sex, and age. Autistic adults had uniquely lower cognitive empathy and a higher “D-score” (reflecting higher systemising than empathy); lower empathy was also observed in autistic children and adolescents. There were sex-by-diagnosis and age-by-diagnosis interactions for empathy in adults, with higher empathy in females and older adults in the non-autistic population only. In this population, higher SA (but not CT) corresponded to lower empathy and a higher D-score; the empathy relationships were observed for SA clusters underlying meta-analytic task-based activations of empathy, both cognitive and affective. Selectively, SA within these empathy clusters, and SA within functional networks spanning the entire cortex, interacted with autism in relation to empathy, differed by sex, and differed by age from childhood to adulthood. (...) | 725/725 | Secondary Analysis | Shared |
| Cross-cohort replicability and generalizability of connectivity-based psychometric prediction patterns | 10.15154/1524254 | An increasing number of studies have investigated the relationships between inter-individual variability in brain regions’ connectivity and behavioral phenotypes, making use of large population neuroimaging datasets. However, the replicability of brain-behavior associations identified by these approaches remains an open question. In this study, we examined the cross-dataset replicability of brain-behavior association patterns for fluid cognition and openness predictions using a previously developed region-wise approach, as well as using a standard whole-brain approach. Overall, we found moderate similarity in patterns for fluid cognition predictions across cohorts, especially in the Human Connectome Project Young Adult, Human Connectome Project Aging, and Enhanced Nathan Kline Institute Rockland Sample cohorts, but low similarity in patterns for openness predictions. In addition, we assessed the generalizability of prediction models in cross-dataset predictions, by training the model in one dataset and testing in another. Making use of the region-wise prediction approach, we showed that first, a moderate extent of generalizability could be achieved with fluid cognition prediction, and that, second, a set of common brain regions related to fluid cognition across cohorts could be identified. Nevertheless, the moderate replicability and generalizability could only be achieved in specific contexts. Thus, we argue that replicability and generalizability in connectivity-based prediction remain limited and deserve greater attention in future studies. | 725/725 | Secondary Analysis | Shared |
| Human Connectome Project-Aging (HCP-A) Release 2.0 | 10.15154/1520707 | The 2.0 release of data from the Human Connectome Project in Aging (healthy participants, ages 36-100+) includes visit 1 (V1) preprocessed structural and functional imaging data, unprocessed V1 imaging data for all modalities (structural, high-res hippocampal T2, resting state fMRI, task fMRI, diffusion, and ASL), and non-imaging demographic and behavioral assessment data for 725 participants. For details of all the measures included in this release and access instructions see the Lifespan HCP-Aging Release 2.0 documentation link below.
| 725/725 | Primary Analysis | Shared |
| Multivariate associations of motor performance, sleep quality, depressive symptoms, and grey matter volume in younger and mid-to-older adults | 10.15154/qq04-sd33 | Motor performance (MP) is essential for maintaining functional independence, particularly in later life. However, the relationship between MP and sleep quality, depressive symptoms, and their underlying brain substrates remains obscure. We employed four samples of younger/mid-to-older adults (n=1,954) from the Human Connectome Project-Young Adult (HCP-YA), HCP-Aging (HCP-A), and enhanced Nathan Kline Institute-Rockland sample (eNKI-RS) to assess the replicability of our findings. Using canonical correlation analyses within a machine learning framework, we investigated the associations of sleep quality, depressive symptoms, and grey matter volume (GMV) with MP. In the combined model of the HCP-YA sample, a canonical variate of better sleep, mild, sub-clinical depressive symptoms, and altered GMV of several cortical (including precentral and fusiform gyrus), thalamus, and cerebellar brain regions was associated with a canonical variate of better MP (r=0.2, SD=0.05). This pattern was conceptually replicated in the young eNKI-RS sample (r=0.25, SD=0.13). In the HCP-A sample, a variate of better sleep quality, fewer depressive symptoms, and increased GMV was associated with a variate of MP (r=0.18, SD=0.1), but these findings did not replicate in the mid-to-older eNKI-RS sample (r=0, SD=0.12). Across all samples, variates of increased GMV were associated with variates of better MP, suggesting potential neuroanatomical underpinnings. We observed age-related variations in the multivariate associations between sleep quality, depressive symptoms, and GMV with MP. | 725/725 | Secondary Analysis | Shared |
| Physical Fitness, Cognition, and Structural Network Efficiency of Brain Connections Across the Lifespan | 10.15154/1520882 | Inadequate levels of exercise is one of the most potent modifiable risk factors for preventing cognitive decline and dementia as we age. Meanwhile, network science-based measures of structural brain network global and local efficiency show promise as robust biomarkers of aging, cognitive decline, and pathological disease progression. Despite this, little to no work has established how maintaining physical activity (PA) and physical fitness might relate to cognition and network efficiency measures across the lifespan. Therefore the purpose of this study was to determine the relationship between (1) PA and fitness and cognition, (2) fitness and network efficiency, and (3) how network efficiency measures relate to cognition. To accomplish this, we analyzed a large cross-sectional data set (n=720; 36-100 years) from the aging human connectome project, which included the Trail Making Task (TMT) A and B, a measure of fitness (2-minute walk test), physical activity (International Physical Activity Questionnaire), and high-resolution diffusion imaging data. Our analysis consisted of employing multiple linear regression while controlling for age, sex, and education. Age was associated with lower global and local brain network efficiency and poorer Trail A & B performance. Meanwhile, fitness, but not physical activity, was related to better Trail A and B performance and fitness, and was positively associated with local and global brain efficiency. Finally, local efficiency was related to better TMT B performance and partially mediated the relationship between fitness and TMT B performance. These results indicate aging may be associated with a shift towards less efficient local and global neural networks and that maintaining physical fitness might protect against age-related cognitive performance deterioration by bolstering structural network efficiency. | 725/725 | Secondary Analysis | Shared |
| Regional neuroanatomic effects on brain age inferred using magnetic resonance imaging and ridge regression | 10.15154/h9fm-ff03 | The biological age of the brain differs from its chronological age (CA) and can be used as biomarker of neural/cognitive disease processes and as predictor of mortality. Brain age (BA) is often estimated from magnetic resonance images (MRIs) using machine learning (ML) that rarely indicates how regional brain features contribute to BA. Leveraging an aggregate training sample of 3 418 healthy controls (HCs), we describe a ridge regression model that quantifies each region’s contribution to BA. After model testing on an independent sample of 651 HCs, we compute the coefficient of partial determination R¯2p for each regional brain volume to quantify its contribution to BA. Model performance is also evaluated using the correlation r between chronological and biological ages, the mean absolute error (MAE) and mean squared error (MSE) of BA estimates. On training data, r=0.92, MSE=70.94 years, MAE=6.57 years, and R¯2=0.81; on test data, r=0.90, MSE=81.96 years, MAE=7.00 years, and R¯2=0.79. The regions whose volumes contribute most to BA are the nucleus accumbens (R¯2p=7.27%), inferior temporal gyrus (R¯2p=4.03%), thalamus (R¯2p=3.61%), brainstem (R¯2p=3.29%), posterior lateral sulcus (R¯2p=3.22%), caudate nucleus (R¯2p=3.05%), orbital gyrus (R¯2p=2.96%), and precentral gyrus (R¯2p=2.80%). Our ridge regression, although outperformed by the most sophisticated ML approaches, identifies the importance and relative contribution of each brain structure to overall BA. Aside from its interpretability and quasi-mechanistic insights, our model can be used to validate future ML approaches for BA estimation. | 725/725 | Primary Analysis | Shared |
| Sleep Quality Moderates Associations between Fitness and Hippocampal and Entorhinal Structure in Middle Aged and Older Adults | 10.15154/71fs-xs37 | As individuals age, the entorhinal cortex (ERC) and hippocampus, crucial structures for memory tend to experience atrophy, and relate to cognitive decline. Simultaneously, lifestyle factors that can be modified, such as exercise and sleep, have been separately shown to slow atrophy and functional decline. Yet, the synergistic impact of fitness and sleep quality on susceptible brain structures in aging adults remains uncertain. This study aimed to examine both independent and interactive relationships between fitness and subjective sleep quality and their influence on ERC thickness and hippocampal volume. We conducted our analysis on data obtained from a sample of 598 middle-aged and older adults from the Human Connectome Lifespan Aging Project. Cardiorespiratory fitness was assessed using the 2-minute walk test (2MWT), while subjective sleep quality was measured with the Pittsburgh Sleep Quality Index (PSQI). High-resolution structural magnetic resonance imaging were used to examine mean ERC thickness and bilateral hippocampal volume. Through multiple linear regression analyses, we explored the moderating effects of subjective sleep quality on the relationship between fitness and brain structure, accounting for variables such as age, sex, education, body mass index, gait speed, and subjective physical activity. Our findings revealed that greater cardiorespiratory fitness, but not subjective sleep quality, was positively associated with bilateral hippocampal volume and ERC thickness. Notably, the positive correlation between fitness and both hippocampal volume and ERC thickness was significantly diminished among individuals reporting the poorest subjective sleep quality. In conclusion, this study underscores the importance of both cardiorespiratory fitness and subjective sleep quality in preserving critical, age-vulnerable brain structures. It suggests that interventions targeting sleep health and exercise should take into consideration the combined effects of sleep and fitness on brain health. | 725/725 | Secondary Analysis | Shared |
| Structural Brain Changes in Emotion Recognition Across the Adult Lifespan | 10.15154/wmt5-k422 | Emotion recognition (ER) declines with increasing age, yet little is known whether this observation bases on structural brain changes conveyed by differential atrophy. To investigate whether age-related ER decline correlates with reduced grey matter (GM) volume in emotion-related brain regions, we conducted a voxel-based morphometry analysis using data of the Human Connectome Project-Aging (N = 238, aged 36 - 87) in which facial ER was tested. We expected to find brain regions that show an additive or super-additive age-related change in GM volume indicating atrophic processes that reduce ER in older adults. The data did not support our hypotheses after correction for multiple comparisons. Exploratory analyses with a threshold of p < .001 (uncorrected), however, suggested that relationships between GM volume and age-related general ER may be widely distributed across the cortex. Yet, small effect sizes imply that only a small fraction of the decline of ER in older adults can be attributed to local GM volume changes in single voxels or their multivariate patterns. | 725/725 | Primary Analysis | Shared |
| Transfer learning for cognitive reserve quantification | 10.15154/1528404 | Cognitive reserve (CR) has been introduced to explain individual differences in susceptibility to cognitive or functional impairment in the presence of age or pathology. We developed a deep learning model to quantify the CR as residual variance in memory performance using the Structural Magnetic Resonance Imaging (sMRI) data from a lifespan healthy cohort. The generalizability of the sMRI-based deep learning model was tested in two independent healthy and Alzheimer's cohorts using transfer learning framework. Structural MRIs were collected from three cohorts: 495 healthy adults (age: 20-80) from RANN, 620 healthy adults (age: 36-100) from lifespan Human Connectome Project Aging (HCPA), and 941 adults (age: 55-92) from Alzheimer's Disease Neuroimaging Initiative (ADNI). Region of interest (ROI)-specific cortical thickness and volume measures were extracted using the Desikan-Killiany Atlas. CR was quantified by residuals which subtract the predicted memory from the true memory. Cascade neural network (CNN) models were used to train RANN dataset for memory prediction. Transfer learning was applied to transfer the T1 imaging-based model from source domain (RANN) to the target domains (HCPA or ADNI). The CNN model trained on the RANN dataset exhibited strong linear correlation between true and predicted memory based on the T1 cortical thickness and volume predictors. In addition, the model generated from healthy lifespan data (RANN) was able to generalize to an independent healthy lifespan data (HCPA) and older demented participants (ADNI) across different scanner types. The estimated CR was correlated with CR proxies such education and IQ across all three datasets. The current findings suggest that the transfer learning approach is an effective way to generalize the residual-based CR estimation. It is applicable to various diseases and may flexibly incorporate different imaging modalities such as fMRI and PET, making it a promising tool for scientific and clinical purposes. | 725/725 | Secondary Analysis | Shared |
| Translating phenotypic prediction models from big to small anatomical MRI data using meta-matching | 10.15154/50bk-b953 | Individualized phenotypic prediction based on structural magnetic resonance imaging (MRI) is an important goal in neuroscience. Prediction performance increases with larger samples, but small-scale datasets with fewer than 200 participants are often unavoidable. We have previously proposed a “meta-matching” framework to translate models trained from large datasets to improve the prediction of new unseen phenotypes in small collection efforts. Meta-matching exploits correlations between phenotypes, yielding large improvement over classical machine learning when applied to prediction models using resting-state functional connectivity as input features. Here, we adapt the two best performing meta-matching variants (“meta-matching finetune” and “meta-matching stacking”) from our previous study to work with T1-weighted MRI data by changing the base neural network architecture to a 3D convolution neural network. We compare the two meta-matching variants with elastic net and classical transfer learning using the UK Biobank (N = 36,461), the Human Connectome Project Young Adults (HCP-YA) dataset (N = 1,017), and the HCP-Aging dataset (N = 656). We find that meta-matching outperforms elastic net and classical transfer learning by a large margin, both when translating models within the same dataset and when translating models across datasets with different MRI scanners, acquisition protocols, and demographics. For example, when translating a UK Biobank model to 100 HCP-YA participants, meta-matching finetune yielded a 136% improvement in variance explained over transfer learning, with an average absolute gain of 2.6% (minimum = –0.9%, maximum = 17.6%) across 35 phenotypes. Overall, our results highlight the versatility of the meta-matching framework. | 725/725 | Secondary Analysis | Shared |
| anatomically interpretable deep learning of brain age captures domain-specific cognitive impairment | 10.15154/qx0r-z707 | The gap between chronological age (CA) and biological brain age, as estimated from magnetic resonance images (MRIs), reflects how individual patterns of neuroanatomic aging deviate from their typical trajectories. MRI-derived brain age (BA) estimates are often obtained using deep learning models that may perform relatively poorly on new data or that lack neuroanatomic interpretability. This study introduces a convolutional neural network (CNN) to estimate BA after training on the MRIs of 4,681 cognitively normal (CN) participants and testing on 1,170 CN participants from an independent sample. BA estimation errors are notably lower than those of previous studies. At both individual and cohort levels, the CNN provides detailed anatomic maps of brain aging patterns that reveal sex dimorphisms and neurocognitive trajectories in adults with mild cognitive impairment (MCI, N = 351) and Alzheimer’s disease (AD, N = 359). In individuals with MCI (54% of whom were diagnosed with dementia within 10.9 y from MRI acquisition), BA is significantly better than CA in capturing dementia symptom severity, functional disability, and executive function. Profiles of sex dimorphism and lateralization in brain aging also map onto patterns of neuroanatomic change that reflect cognitive decline. Significant associations between BA and neurocognitive measures suggest that the proposed framework can map, systematically, the relationship between aging-related neuroanatomy changes in CN individuals and in participants with MCI or AD. Early identification of such neuroanatomy changes can help to screen individuals according to their AD risk.
| 725/725 | Primary Analysis | Shared |
| Metabolism modulates network synchrony in the aging brain | 10.15154/1526427 | Brain aging is associated with hypometabolism and global changes in functional connectivity. Using functional MRI (fMRI), we show that network synchrony, a collective property of brain activity, decreases with age. Applying quantitative methods from statistical physics, we provide a generative (Ising) model for these changes as a function of the average communication strength between brain regions. We find that older brains are closer to a critical point of this communication strength, in which even small changes in metabolism lead to abrupt changes in network synchrony. Finally, by experimentally modulating metabolic activity in younger adults, we show how metabolism alone—independent of other changes associated with aging—can provide a plausible candidate mechanism for marked reorganization of brain network topology. | 719/719 | Secondary Analysis | Shared |
| The interaction between age and sleep quality impacts cognition | 10.15154/1528398 | Background: The association between sleep quality and cognition is widely established but the role of the aging process on this relationship is largely unknown.
Purpose: To examine how age impacts the sleep-cognition relationship and determine critical age ranges when sleep is most strongly associated with cognition. This investigation could help identify individuals at risk for sleep-related cognitive impairment.
Methods: Sample included 711 individuals (59.66 ± 14.91, 55.7 % female) who were assessed for sleep quality using the Pittsburgh Sleep Quality Index (PSQI) and cognition as part of the Human Connectome Project Aging (HCP-A).
Results: There was a significant interaction term between the PSQI and non-linear age term (age2) on Trail Making Test B (TMT-B) (p = 0.02) and NIH Toolbox (TB) crystallized cognition (p = 0.02), with critical age ranges at ages 50-75 for TMT-B and ages 66-70 for crystallized cognition.
Conclusions: The relationship between sleep quality and cognitive performance may be modified by age. Furthermore, middle-age to early older adulthood may be the most vulnerable to sleep-related cognitive impairment.
| 712/712 | Secondary Analysis | Shared |
| Human Connectome Project-Aging (HCP-A) Release 1.0 | 10.15154/1503533 | Initial release of data from the Human Connectome Project in Aging (ages 35-100+). Release includes basic demographic data (sex, age, race/ethnicity, handedness) and unprocessed imaging data for all modalities (structural, high-res hippocampal T2, resting state fMRI, task fMRI, diffusion, and ASL) for 689 subjects and preprocessed structural imaging data for 128 subjects. Full release documentation available at: https://www.humanconnectome.org/study/hcp-lifespan-aging/documentation
| 689/689 | Primary Analysis | Shared |
| Revealing the spatial pattern of brain hemodynamic sensitivity to healthy aging through sparse DCM | 10.15154/hk8k-pm84 | Age-related changes in the BOLD response could reflect neuro-vascular coupling modifications rather than simply impairments in neural functioning. In this study, we propose the use of a generative dynamic causal model (DCM) to decouple neuronal and vascular factors in the BOLD signal, with the aim of characterizing the whole-brain spatial pattern of hemodynamic sensitivity to healthy aging, as well as to test the role of hemodynamic features as independent predictors in an age-classification model.
In this view, DCM was applied to the resting-state fMRI data of a cohort of 126 healthy individuals in a wide age range, providing reliable estimates of the hemodynamic response function (HRF) for each subject and each region of interest. Then, some features characterizing each HRF curve were extracted and used to fit a multivariate logistic regression model to predict the age class of each individual. Ultimately, we tested the final predictive model on an independent dataset of 338 healthy subjects selected from the Human Connectome Project Aging (HCP-A) and Development (HCP-D) cohorts. Our results entail the spatial heterogeneity of the age effects on the hemodynamic component, since its impact resulted to be strongly region- and population-specific, discouraging any space-invariant corrective procedures that attempt to correct for vascular factors when carrying out functional studies involving groups with different ages. Moreover, we demonstrated that a strong interaction exists between some specific hemodynamic features and age, further supporting the essential role of the hemodynamic factor as independent predictor of biological aging, rather than a simple confounding variable.
| 489/641 | Secondary Analysis | Shared |
| Sex-differential markers of psychiatric risk and treatment response based on premature aging of functional brain network dynamics and peripheral physiology | 10.15154/9wfk-qa89 | Background
Aging is a multilevel process of gradual decline that predicts morbidity and mortality. Independent investigations have implicated senescence of brain and peripheral physiology in psychiatric risk, but it is unclear whether these effects stem from unique or shared mechanisms.
Methods
To address this question, we analyzed clinical, blood chemistry and resting state functional neuroimaging data in a healthy aging cohort (N= 427; age 36-100 years) and two disorder-specific samples encompassing patients with early psychosis (100 patients, 16-35 years) and major depressive disorder (MDD) (104 patients, 20-76 years).
Results
We identified sex-dependent coupling between blood chemistry markers of metabolic senescence (i.e., homeostatic dysregulation), functional brain network aging, and psychiatric risk. In females, premature aging of frontoparietal and somatomotor networks was linked to greater homeostatic dysregulation. It also predicted the severity and treatment resistance of mood symptoms (depression/anxiety [all three samples], anhedonia [MDD]) and social withdrawal/behavioral inhibition (avoidant personality disorder [healthy aging]; negative symptoms [early psychosis]). In males, premature aging of the default mode, cingulo-opercular, and visual networks was linked to reduced homeostatic dysregulation and predicted severity and treatment resistance of symptoms relevant to hostility/aggression (antisocial personality disorder [healthy aging]; mania/positive symptoms [early psychosis]), impaired thought processes (early psychosis, MDD) and somatic problems (healthy aging, MDD).
Conclusions
Our findings identify sexually dimorphic relationships between brain dynamics, peripheral physiology, and risk for psychiatric illness, suggesting that the specificity of putative risk biomarkers and precision therapeutics may be improved by considering sex and other relevant personal characteristics.
| 427/631 | Secondary Analysis | Shared |
| Neurofind: Using deep learning to make individualised inferences in brain-based disorders | 10.15154/rad1-0h66 | Within precision psychiatry, there is a growing interest in normative models given their ability to parse het-erogeneity. Although these models are intuitive and informative, their development typically requires significant technical expertise and resources, posing a challenge for many researchers. Here, we present Neurofind, a new freely available research tool that bridges this gap by wrapping sound and previously tested methods on data harmonisation and advanced normative models into a web-based platform that requires minimal input from the user. Neurofind takes structural magnetic resonance imaging (MRI) images as input and generates two key metrics using independent normative models: 1) Outlier index, a deviation score from the normative brain morphology and 2) Brain age, the predicted age based on an individual’s brain morphometry. The tool was trained on 3362 images of healthy controls aged 20 to 80, sourced from publicly available datasets. The grey matter volume of 101 cortical and subcortical regions was extracted and mod-elled with a generative deep autoencoder for the Outlier index model and a support vector regression for the Brain age model. To demonstrate the potential applications of Neurofind, we applied the tool to 725 images from three independent datasets of patients diagnosed with Alzheimer’s disease and schizophrenia. In Alzheimer’s disease, 55.2% of patients were at high risk of being outliers from the normative brain morphology, mostly driven by larger deviations in the ventricles and limbic structures. Patients were also homogeneous in how they deviated from the norm. Conversely, only 30.1% of schizophrenia patients were at high risk of being outliers, mostly due to deviations in the hippocampus, striatum and pallidum, and patients were highly heterogeneous. Both groups showed signs of accelerated brain aging. Neurofind can be accessed via the website www.neurofind.ai and is accompanied by a comprehensive how-to guide, which outlines the tool’s usage in four simple steps. | 623/623 | Secondary Analysis | Shared |
| Educational quality may be a closer correlate of cardiometabolic health than educational attainment | 10.15154/1528124 | Educational quality may be a closer correlate of physical health than more commonly used measures of educational attainment (e.g., years in school). We examined whether a widely-used performance-based measure of educational quality is more closely associated with cardiometabolic health than educational attainment (highest level of education completed), and whether perceived control (smaller sample only), executive functioning (both samples), and health literacy (smaller sample only) link educational quality to cardiometabolic health. In two samples (N=98 and N=586) collected from different regions of the US, educational quality was associated with cardiometabolic health above and beyond educational attainment, other demographic factors (age, ethnoracial category, sex), and fluid intelligence. Counter to expectations, neither perceived control, executive function, nor health literacy significantly mediated the association between educational quality and cardiometabolic health. Findings add to the growing literature suggesting that current operationalizations of the construct of education likely underestimate the association between education and multiple forms of health. To the extent that educational programs may have been overlooked based on the apparent size of associations with outcomes, such actions may have been premature. | 586/586 | Secondary Analysis | Shared |
| Intermediately Synchronised Brain States optimise trade-off between Subject Identifiability and Predictive Capacity | 10.15154/1527952 | Functional connectivity (FC) refers to the statistical dependencies between activity of distinct brain
areas. To study temporal fluctuations in FC within the duration of a functional magnetic resonance
imaging (fMRI) scanning session, researchers have proposed the computation of an edge time series
(ETS) and their derivatives. Evidence suggests that FC is driven by a few time points of high-amplitude co-
fluctuation (HACF) in the ETS, which may also contribute disproportionately to interindividual differences.
However, it remains unclear to what degree different time points actually contribute to brain-behaviour
associations. Here, we systematically evaluate this question by assessing the predictive utility of FC
estimates at different levels of co-fluctuation using machine learning (ML) approaches. We demonstrate
that time points of lower and intermediate co-fluctuation levels provide overall highest subject specificity
as well as highest predictive capacity of individual-level phenotypes.
| 558/558 | Secondary Analysis | Shared |
| Brain Topology Underlying Executive Functions Across the Lifespan: Focus on the Default Mode Network | 10.15154/pvjz-h270 | While traditional neuroimaging approaches to the study of executive functions (EFs) have typically employed task-evoked paradigms, resting state studies are gaining popularity as a tool for investigating inter-individual variability in the functional connectome and its relationship to cognitive performance outside of the scanner. Using resting state functional magnetic resonance imaging data from the Human Connectome Project Lifespan database, the present study capitalised on graph theory to chart cross-sectional variations in the intrinsic functional organisation of the frontoparietal (FPN) and the default mode (DMN) networks in 500 healthy individuals (from 10 to 100 years of age), to investigate the neural underpinnings of EFs across the lifespan. Topological properties of both the FPN and DMN were predictive of EF performance, but not of a control task of picture naming, providing specificity in support for a tight link between neuro-functional and cognitive-behavioural efficiency within the EF domain. The topological organisation of the DMN, however, appeared more sensitive to age-related changes relative to that of the FPN. Because the DMN matures earlier in life than the FPN, it is more susceptible to neurodegenerative changes. Moreover, because its activity is stronger in conditions of resting state, the DMN might be easier to measure in noncompliant populations and in those at the extremes of the life-span curve, namely very young or elder participants. Here, we argue that the study of its functional architecture in relation to higher order cognition across the lifespan might, thus, be of greater interest compared with what has been traditionally thought. | 407/500 | Secondary Analysis | Shared |
| Identifying and characterizing cognitive profiles in midlife females: A latent profile analysis | 10.15154/xrsr-6c85 | Females are at greater risk of developing Alzheimer’s disease (AD) than men. The menopause transition, which involves a neuroendocrine shift, is a potential contributor to this sex difference. Multiple estrogen-regulated systems (i.e., circadian rhythms) are disrupted during this transition which may affect cognitive functioning (Barha & Liu-Ambrose, 2020), most notably verbal learning and memory. Midlife females are chronically understudied, and little is known about how individual factors (i.e., sleep, physical activity (PA), stress, depressive symptoms) may relate to cognitive functioning across midlife for females, a period marked by the menopausal transition. Utilizing data from the Human Connectome Aging project (HCP-A), the current study will examine whether distinct cognitive profiles determined by performance-based tasks relate to emotional and physical functioning among a sample of middle-aged females.
Late reproductive, perimenopause, and postmenopausal females (ages 40 to 60) from the HCP-A were included (n =202, M age = 50.5, SD = 6.2). Demographic information, sleep problems (Pittsburgh Sleep Quality Index), PA (International Physical Activity Questionnaire), stress (Distress subscale of the Perceived Stress Scale), depressive symptoms (NIH toolbox Emotion Module) were assessed with surveys, and participants completed several performance-based tasks including: global cognitive function (MoCA), dimensional change card sort (DCCS), flanker, pattern recognition, working memory (WM), picture sequencing, receptive language task, Trail Making Test B (TMT-B), and Rey Auditory Verbal Learning (RAVLT) tasks. Using latent profile analysis (LPA), cognitive profiles were identified via performance-based cognitive tasks. Emergent profiles were characterized in terms of demographic information, psychological, and behavioral factors.
Fit indices indicated that a three-class solution fit the sample best: below average to low average performance across domains (Class 1, n=22), average performance across domains (Class 2, n= 72) and low average to average performance across domains (Class 3, n= 76). There was no significant multivariate effect of cognitive profile on psychological and behavioral factors, (p= .14), after controlling for age and education. Univariate analyses revealed significant differences between classes based on depressive symptoms, F(2,52.7) = 3.69, p = .027, η^2= .043) such that females in Class 1 reported higher levels of symptoms than both Class 2 and 3. There was no difference between Class 2 and 3 regarding depressive symptoms, and contrary to hypotheses, no difference in PA, stress, or sleep problems were observed between any classes.
Results suggest three distinct cognitive profiles exist in this analytic sample. After controlling for age and education, only depressive symptoms significantly differed between cognitive profiles. The class characterized by low to below-average cognitive performance demonstrated higher levels of reported depressive symptoms as compared to other classes. These findings provide preliminary evidence that middle-aged females who perform worse on cognitive tasks may be experiencing heightened depressive symptoms, which are known to worsen in the perimenopause. Future research should explore more psychological and behavioral factors and whether emerging associations are moderated by menopausal stage.
| 332/332 | Secondary Analysis | Shared |
| Neurocognitive Differences Between Lifestyle Profiles of Women Across the Menopausal Transition | 10.15154/1528445 | Women are at greater risk of developing Alzheimer’s disease (AD) than men. The menopausal transition, which involves a neuroendocrine shift for women, is a potential contributor to this sex difference. Multiple estrogen-regulated systems (i.e., circadian rhythms) are disrupted during this transition which affects cognitive functioning (Barha & Liu-Ambrose, 2020), most notably verbal learning and memory. Little is known about how lifestyle factors (i.e., sleep, physical activity (PA), stress) may promote neurocognitive functioning across this transition (Maki & Weber, 2021). Utilizing data from the Human Connectome Aging project (HCP-A), the current study will examine whether distinct lifestyle profiles including sleep, PA, and stress relate to multiple domains of cognitive performance among a sample of perimenopausal/menopausal women.
Perimenopausal/menopausal women (ages 45 to 65) from the HCP-A were included (n =150, M age = 54.6, SD = 5.5). Demographic information, menopausal status, sleep problems (Pittsburgh Sleep Quality Index), PA (International Physical Activity Questionnaire), stress (Distress subscale of the Perceived Stress Scale) were assessed with surveys, and participants completed several lab-based tasks including: dimensional change card sort (DCCS), flanker, pattern recognition, working memory (WM), picture sequencing, oral reading, Trails Making Test A and B (TMT), and Rey Auditory Verbal Learning (RAVLT) tasks. Using latent profile analysis (LPA), lifestyle profiles were identified via sleep problems, PA, and stress levels. A MANOVA compared cognitive performance between these lifestyle profiles, above and beyond age and education status.
Fit indices indicated that a three-class solution fit the sample best: high PA, low stress and sleep problems (Class 1, n=38), high PA, stress, and sleep problems (Class 2, n= 17) and low PA, high stress and sleep problems (Class 3, n= 95) which were not significantly different based on age or menopausal status (p>0.05). A significant multivariate effect of age and education on cognitive performance (p<.001) emerged. There was a significant multivariate effect of lifestyle profile on cognitive performance, F (18, 260) = 1.73, p=.034, η^2= .11, after controlling for age and education. Univariate analyses determined that certain lifestyle profiles were associated with better performance on all cognitive tasks except verbal memory. Contrary to expectation, Class 3 performed better on TMT A & B, DCCS, flanker, WM, and pattern recognition tasks as compared to Class 1. Class 3 performed better on reading and picture sequencing tasks than Class 2. There was no difference in performance between Class 1 and 2.
Results suggest three distinct lifestyle profiles exist in this analytic sample. After controlling for age and education, cognitive performance on all tasks except for verbal memory significantly differed between lifestyle profiles. The profile characterized by low PA and high stress and sleep problems demonstrated superior performance as compared to other classes. These findings provide preliminary evidence that women who have high levels of stress and sleep problems with low PA are performing better on cognitive tasks, but replication of these findings utilizing longitudinal designs are needed.
| 332/332 | Secondary Analysis | Shared |
| Psychological Resilience and Neurodegenerative Risk: A Connectomics-Transcriptomics Investigation in Healthy Adolescent and Middle-Aged Females | 10.15154/1526352 | Adverse life events can inflict substantial long-term damage, which, paradoxically, has been posited to stem from initially adaptative responses to the challenges encountered in one’s environment. Thus, identification of the mechanisms linking resilience against recent stressors to longer-term psychological vulnerability is key to understanding optimal functioning across multiple timescales. To address this issue, our study tested the relevance of neuro-reproductive maturation and senescence, respectively, to both resilience and longer-term risk for pathologies characterised by accelerated brain aging, specifically, Alzheimer’s Disease (AD). Graph theoretical and partial least squares analyses were conducted on multimodal imaging, reported biological aging and recent adverse experience data from the Lifespan Human Connectome Project (HCP). Availability of reproductive maturation/senescence measures restricted our investigation to adolescent (N =178) and middle-aged (N=146) females. Psychological resilience was linked to age-specific brain senescence patterns suggestive of precocious functional development of somatomotor and control-relevant networks (adolescence) and earlier aging of default mode and salience/ventral attention systems (middle adulthood). Biological aging showed complementary associations with the neural patterns relevant to resilience in adolescence (positive relationship) versus middle-age (negative relationship). Transcriptomic and expression quantitative trait locus data analyses linked the neural aging patterns correlated with psychological resilience in middle adulthood to gene expression patterns suggestive of increased AD risk. Our results imply a partially antagonistic relationship between resilience against proximal stressors and longer-term psychological adjustment in later life. They thus underscore the importance of fine-tuning extant views on successful coping by considering the multiple timescales across which age-specific processes may unfold. | 146/324 | Secondary Analysis | Shared |
