There are many active research projects accessing and applying shared ADNI data. Use the search above to find specific research focuses on the active ADNI investigations. This information is requested annually as a requirement for data access.
| Principal Investigator | |
| Principal Investigator's Name: | Makenna McGill |
| Institution: | The University of Texas at Austin |
| Department: | Psychology |
| Country: | |
| Proposed Analysis: | Examining Accelerated Brain Aging Associated with Service-related Traumatic Brain Injury Using Multimodal Neuroimaging Emerging evidence suggests that a traumatic brain injury (TBI) sustained during early adulthood may influence aging trajectories and increase risk of developing dementia. Although the precise mechanisms underlying this relationship are poorly understood, remote TBI may accelerate normal age-related changes such as cerebral white matter (WM) alterations and cortical gray matter (GM) thinning, resulting in cognitive decline. Brain age models of these diffusion and structural MRI metrics offer a powerful method of identifying accelerated brain aging in clinical samples. Brain-predicted age difference (brain-PAD) can be calculated as the discrepancy between an individual’s model-predicted age and their chronological age, with higher brain-PAD indicating an acceleration of the normal aging trajectory. WM integrity and cortical thinning have been independently studied in Veterans with a history of TBI, however, these studies were not focused on older Veterans. It therefore remains unclear how these processes might jointly contribute to accelerated brain aging in older Veterans who sustained TBIs decades prior. Given the prevalence of TBI in U.S. military members and the growing population of Veterans ages 65 and older, the current study sought to investigate the effects of service-related TBI on brain aging trajectories using a multimodality neuroimaging model of brain age, and to explore how brain age is related to Veterans’ cognitive, psychological, and vascular health. Aim 1: Determine if brain-predicted age difference (brain-PAD) is moderated by remote TBI in older Veterans by comparing brain age in older Veterans with and without a history of TBI. We hypothesize that Veterans with remote TBI will have greater brain-PAD. Aim 2: Examine which neuroimaging metrics contribute the greatest variance to the brain-PAD. We hypothesize that the diffusion metrics in specific WM tracts will contribute the greatest variance. Aim 3: Explore if brain-PAD is associated with cognitive functioning, psychological functioning, and/or vascular burden. We hypothesize that greater brain-PAD will be associated with poorer cognitive functioning, poorer psychological functioning, and higher vascular burden. Research Plan: The proposed study will use data obtained from the ADNI database to train the brain age model, which will then be tested on data previously obtained from the DOD-ADNI database. ADNI Sample: The training set will consist of 200 cognitively unimpaired, male civilians ages 55-90 without a history of TBI. DOD-ADNI Sample: The test set consists of 195 male Vietnam War Veterans ages 60-80 (mean age=69.3). 102 Veterans met criteria for TBI by sustaining a head injury that resulted in a loss of consciousness, alteration of consciousness, or post-traumatic amnesia. Veterans were excluded if they had a history of alcohol or substance abuse within the past five years, mild cognitive impairment/dementia, or sustained a TBI within 12 months of data collection. Cognitive Measures: Neuropsychological assessment scores will be converted to sample-specific z-scores in cognitive domains of executive functioning (Trail Making Test Parts A and B) and verbal memory/learning (Logical Memory I and II; Rey Auditory Verbal Learning Test). Psychological Measures: Severity of current and lifetime posttraumatic stress disorder (PTSD) will be determined by the Clinician Administered PTSD Scale. The Geriatric Depression Scale will be used to measure depression. Vascular Measures: A vascular burden score (0-5) will be calculated based on the presence of 1) hypertension, 2) diabetes, 3) cardiovascular disease, 4) atrial fibrillation, and 5) transient ischemic attack or minor stroke. Other Measures: Covariates will include intracranial volume (ICV), injury severity, time since last injury, years of education, and preinjury cognitive function (Armed Forces Qualification Test). Diffusion Metrics: Diffusion MRI data will be used to calculate averages of fractional anisotropy (FA, the degree of directionality of water diffusion) and mean diffusivity (MD, which describes the overall water diffusion) for 28 WM pathways of interest (POIs). Structural Metrics: Structural MRI data will be used to calculate cortical thickness for 148 regions of interest (ROIs), and to calculate 11 global brain measures (total cortical and subcortical volumes, mean cortical thickness, total cortical surface area, total cortical and cerebellar GM volumes, total cortical and cerebellar WM volumes, brain stem volume, corpus callosum volume, and white matter hypointensities). Brain Age Model: Mean FA/MD values for each POI, cortical thickness values for each ROI, global brain measures, and covariates will be used to predict brain age via relevance vector regression, a machine learning model that applies regularization through a Gaussian prior. We will use 10-fold cross validation to examine model generalizability and mean absolute error to quantify model performance. The trained model with then be applied to the DOD-ADNI dataset of Veterans with and without a history of TBI to determine if there are between-group differences in brain-PAD (calculated as predicted age minus chronological age). Multiple linear regression will be used to determine if brain-PAD is associated with cognitive functioning, psychological functioning, and/or vascular burden. Implications: The ADNI and DOD-ADNI datasets offer the unique opportunity to compare the brains of older adults with and without a history of TBI. The remoteness of the injuries allows us to examine a very important question of how injuries sustained during early adulthood may manifest decades later and interact with brain aging. As our Veterans age, it is important to understand how injuries experienced while serving will influence cognitive aging and therefore their ability to function independently so that we might inform targeted prevention and intervention efforts. |
| Additional Investigators | |
| Investigator's Name: | Grace Jumonville |
| Proposed Analysis: | Examining Accelerated Brain Aging Associated with Service-related Traumatic Brain Injury Using Multimodal Neuroimaging Emerging evidence suggests that a traumatic brain injury (TBI) sustained during early adulthood may influence aging trajectories and increase risk of developing dementia. Although the precise mechanisms underlying this relationship are poorly understood, remote TBI may accelerate normal age-related changes such as cerebral white matter (WM) alterations and cortical gray matter (GM) thinning, resulting in cognitive decline. Brain age models of these diffusion and structural MRI metrics offer a powerful method of identifying accelerated brain aging in clinical samples. Brain-predicted age difference (brain-PAD) can be calculated as the discrepancy between an individual’s model-predicted age and their chronological age, with higher brain-PAD indicating an acceleration of the normal aging trajectory. WM integrity and cortical thinning have been independently studied in Veterans with a history of TBI, however, these studies were not focused on older Veterans. It therefore remains unclear how these processes might jointly contribute to accelerated brain aging in older Veterans who sustained TBIs decades prior. Given the prevalence of TBI in U.S. military members and the growing population of Veterans ages 65 and older, the current study sought to investigate the effects of service-related TBI on brain aging trajectories using a multimodality neuroimaging model of brain age, and to explore how brain age is related to Veterans’ cognitive, psychological, and vascular health. Aim 1: Determine if brain-predicted age difference (brain-PAD) is moderated by remote TBI in older Veterans by comparing brain age in older Veterans with and without a history of TBI. We hypothesize that Veterans with remote TBI will have greater brain-PAD. Aim 2: Examine which neuroimaging metrics contribute the greatest variance to the brain-PAD. We hypothesize that the diffusion metrics in specific WM tracts will contribute the greatest variance. Aim 3: Explore if brain-PAD is associated with cognitive functioning, psychological functioning, and/or vascular burden. We hypothesize that greater brain-PAD will be associated with poorer cognitive functioning, poorer psychological functioning, and higher vascular burden. Research Plan: The proposed study will use data obtained from the ADNI database to train the brain age model, which will then be tested on data previously obtained from the DOD-ADNI database. ADNI Sample: The training set will consist of 200 cognitively unimpaired, male civilians ages 55-90 without a history of TBI. DOD-ADNI Sample: The test set consists of 195 male Vietnam War Veterans ages 60-80 (mean age=69.3). 102 Veterans met criteria for TBI by sustaining a head injury that resulted in a loss of consciousness, alteration of consciousness, or post-traumatic amnesia. Veterans were excluded if they had a history of alcohol or substance abuse within the past five years, mild cognitive impairment/dementia, or sustained a TBI within 12 months of data collection. Cognitive Measures: Neuropsychological assessment scores will be converted to sample-specific z-scores in cognitive domains of executive functioning (Trail Making Test Parts A and B) and verbal memory/learning (Logical Memory I and II; Rey Auditory Verbal Learning Test). Psychological Measures: Severity of current and lifetime posttraumatic stress disorder (PTSD) will be determined by the Clinician Administered PTSD Scale. The Geriatric Depression Scale will be used to measure depression. Vascular Measures: A vascular burden score (0-5) will be calculated based on the presence of 1) hypertension, 2) diabetes, 3) cardiovascular disease, 4) atrial fibrillation, and 5) transient ischemic attack or minor stroke. Other Measures: Covariates will include intracranial volume (ICV), injury severity, time since last injury, years of education, and preinjury cognitive function (Armed Forces Qualification Test). Diffusion Metrics: Diffusion MRI data will be used to calculate averages of fractional anisotropy (FA, the degree of directionality of water diffusion) and mean diffusivity (MD, which describes the overall water diffusion) for 28 WM pathways of interest (POIs). Structural Metrics: Structural MRI data will be used to calculate cortical thickness for 148 regions of interest (ROIs), and to calculate 11 global brain measures (total cortical and subcortical volumes, mean cortical thickness, total cortical surface area, total cortical and cerebellar GM volumes, total cortical and cerebellar WM volumes, brain stem volume, corpus callosum volume, and white matter hypointensities). Brain Age Model: Mean FA/MD values for each POI, cortical thickness values for each ROI, global brain measures, and covariates will be used to predict brain age via relevance vector regression, a machine learning model that applies regularization through a Gaussian prior. We will use 10-fold cross validation to examine model generalizability and mean absolute error to quantify model performance. The trained model with then be applied to the DOD-ADNI dataset of Veterans with and without a history of TBI to determine if there are between-group differences in brain-PAD (calculated as predicted age minus chronological age). Multiple linear regression will be used to determine if brain-PAD is associated with cognitive functioning, psychological functioning, and/or vascular burden. Implications: The ADNI and DOD-ADNI datasets offer the unique opportunity to compare the brains of older adults with and without a history of TBI. The remoteness of the injuries allows us to examine a very important question of how injuries sustained during early adulthood may manifest decades later and interact with brain aging. As our Veterans age, it is important to understand how injuries experienced while serving will influence cognitive aging and therefore their ability to function independently so that we might inform targeted prevention and intervention efforts. |
| Investigator's Name: | Alexandra Clark |
| Proposed Analysis: | Examining Accelerated Brain Aging Associated with Service-related Traumatic Brain Injury Using Multimodal Neuroimaging Emerging evidence suggests that a traumatic brain injury (TBI) sustained during early adulthood may influence aging trajectories and increase risk of developing dementia. Although the precise mechanisms underlying this relationship are poorly understood, remote TBI may accelerate normal age-related changes such as cerebral white matter (WM) alterations and cortical gray matter (GM) thinning, resulting in cognitive decline. Brain age models of these diffusion and structural MRI metrics offer a powerful method of identifying accelerated brain aging in clinical samples. Brain-predicted age difference (brain-PAD) can be calculated as the discrepancy between an individual’s model-predicted age and their chronological age, with higher brain-PAD indicating an acceleration of the normal aging trajectory. WM integrity and cortical thinning have been independently studied in Veterans with a history of TBI, however, these studies were not focused on older Veterans. It therefore remains unclear how these processes might jointly contribute to accelerated brain aging in older Veterans who sustained TBIs decades prior. Given the prevalence of TBI in U.S. military members and the growing population of Veterans ages 65 and older, the current study sought to investigate the effects of service-related TBI on brain aging trajectories using a multimodality neuroimaging model of brain age, and to explore how brain age is related to Veterans’ cognitive, psychological, and vascular health. Aim 1: Determine if brain-predicted age difference (brain-PAD) is moderated by remote TBI in older Veterans by comparing brain age in older Veterans with and without a history of TBI. We hypothesize that Veterans with remote TBI will have greater brain-PAD. Aim 2: Examine which neuroimaging metrics contribute the greatest variance to the brain-PAD. We hypothesize that the diffusion metrics in specific WM tracts will contribute the greatest variance. Aim 3: Explore if brain-PAD is associated with cognitive functioning, psychological functioning, and/or vascular burden. We hypothesize that greater brain-PAD will be associated with poorer cognitive functioning, poorer psychological functioning, and higher vascular burden. Research Plan: The proposed study will use data obtained from the ADNI database to train the brain age model, which will then be tested on data previously obtained from the DOD-ADNI database. ADNI Sample: The training set will consist of 200 cognitively unimpaired, male civilians ages 55-90 without a history of TBI. DOD-ADNI Sample: The test set consists of 195 male Vietnam War Veterans ages 60-80 (mean age=69.3). 102 Veterans met criteria for TBI by sustaining a head injury that resulted in a loss of consciousness, alteration of consciousness, or post-traumatic amnesia. Veterans were excluded if they had a history of alcohol or substance abuse within the past five years, mild cognitive impairment/dementia, or sustained a TBI within 12 months of data collection. Cognitive Measures: Neuropsychological assessment scores will be converted to sample-specific z-scores in cognitive domains of executive functioning (Trail Making Test Parts A and B) and verbal memory/learning (Logical Memory I and II; Rey Auditory Verbal Learning Test). Psychological Measures: Severity of current and lifetime posttraumatic stress disorder (PTSD) will be determined by the Clinician Administered PTSD Scale. The Geriatric Depression Scale will be used to measure depression. Vascular Measures: A vascular burden score (0-5) will be calculated based on the presence of 1) hypertension, 2) diabetes, 3) cardiovascular disease, 4) atrial fibrillation, and 5) transient ischemic attack or minor stroke. Other Measures: Covariates will include intracranial volume (ICV), injury severity, time since last injury, years of education, and preinjury cognitive function (Armed Forces Qualification Test). Diffusion Metrics: Diffusion MRI data will be used to calculate averages of fractional anisotropy (FA, the degree of directionality of water diffusion) and mean diffusivity (MD, which describes the overall water diffusion) for 28 WM pathways of interest (POIs). Structural Metrics: Structural MRI data will be used to calculate cortical thickness for 148 regions of interest (ROIs), and to calculate 11 global brain measures (total cortical and subcortical volumes, mean cortical thickness, total cortical surface area, total cortical and cerebellar GM volumes, total cortical and cerebellar WM volumes, brain stem volume, corpus callosum volume, and white matter hypointensities). Brain Age Model: Mean FA/MD values for each POI, cortical thickness values for each ROI, global brain measures, and covariates will be used to predict brain age via relevance vector regression, a machine learning model that applies regularization through a Gaussian prior. We will use 10-fold cross validation to examine model generalizability and mean absolute error to quantify model performance. The trained model with then be applied to the DOD-ADNI dataset of Veterans with and without a history of TBI to determine if there are between-group differences in brain-PAD (calculated as predicted age minus chronological age). Multiple linear regression will be used to determine if brain-PAD is associated with cognitive functioning, psychological functioning, and/or vascular burden. Implications: The ADNI and DOD-ADNI datasets offer the unique opportunity to compare the brains of older adults with and without a history of TBI. The remoteness of the injuries allows us to examine a very important question of how injuries sustained during early adulthood may manifest decades later and interact with brain aging. As our Veterans age, it is important to understand how injuries experienced while serving will influence cognitive aging and therefore their ability to function independently so that we might inform targeted prevention and intervention efforts. |
| Investigator's Name: | David Schnyer |
| Proposed Analysis: | Examining Accelerated Brain Aging Associated with Service-related Traumatic Brain Injury Using Multimodal Neuroimaging Emerging evidence suggests that a traumatic brain injury (TBI) sustained during early adulthood may influence aging trajectories and increase risk of developing dementia. Although the precise mechanisms underlying this relationship are poorly understood, remote TBI may accelerate normal age-related changes such as cerebral white matter (WM) alterations and cortical gray matter (GM) thinning, resulting in cognitive decline. Brain age models of these diffusion and structural MRI metrics offer a powerful method of identifying accelerated brain aging in clinical samples. Brain-predicted age difference (brain-PAD) can be calculated as the discrepancy between an individual’s model-predicted age and their chronological age, with higher brain-PAD indicating an acceleration of the normal aging trajectory. WM integrity and cortical thinning have been independently studied in Veterans with a history of TBI, however, these studies were not focused on older Veterans. It therefore remains unclear how these processes might jointly contribute to accelerated brain aging in older Veterans who sustained TBIs decades prior. Given the prevalence of TBI in U.S. military members and the growing population of Veterans ages 65 and older, the current study sought to investigate the effects of service-related TBI on brain aging trajectories using a multimodality neuroimaging model of brain age, and to explore how brain age is related to Veterans’ cognitive, psychological, and vascular health. Aim 1: Determine if brain-predicted age difference (brain-PAD) is moderated by remote TBI in older Veterans by comparing brain age in older Veterans with and without a history of TBI. We hypothesize that Veterans with remote TBI will have greater brain-PAD. Aim 2: Examine which neuroimaging metrics contribute the greatest variance to the brain-PAD. We hypothesize that the diffusion metrics in specific WM tracts will contribute the greatest variance. Aim 3: Explore if brain-PAD is associated with cognitive functioning, psychological functioning, and/or vascular burden. We hypothesize that greater brain-PAD will be associated with poorer cognitive functioning, poorer psychological functioning, and higher vascular burden. Research Plan: The proposed study will use data obtained from the ADNI database to train the brain age model, which will then be tested on data previously obtained from the DOD-ADNI database. ADNI Sample: The training set will consist of 200 cognitively unimpaired, male civilians ages 55-90 without a history of TBI. DOD-ADNI Sample: The test set consists of 195 male Vietnam War Veterans ages 60-80 (mean age=69.3). 102 Veterans met criteria for TBI by sustaining a head injury that resulted in a loss of consciousness, alteration of consciousness, or post-traumatic amnesia. Veterans were excluded if they had a history of alcohol or substance abuse within the past five years, mild cognitive impairment/dementia, or sustained a TBI within 12 months of data collection. Cognitive Measures: Neuropsychological assessment scores will be converted to sample-specific z-scores in cognitive domains of executive functioning (Trail Making Test Parts A and B) and verbal memory/learning (Logical Memory I and II; Rey Auditory Verbal Learning Test). Psychological Measures: Severity of current and lifetime posttraumatic stress disorder (PTSD) will be determined by the Clinician Administered PTSD Scale. The Geriatric Depression Scale will be used to measure depression. Vascular Measures: A vascular burden score (0-5) will be calculated based on the presence of 1) hypertension, 2) diabetes, 3) cardiovascular disease, 4) atrial fibrillation, and 5) transient ischemic attack or minor stroke. Other Measures: Covariates will include intracranial volume (ICV), injury severity, time since last injury, years of education, and preinjury cognitive function (Armed Forces Qualification Test). Diffusion Metrics: Diffusion MRI data will be used to calculate averages of fractional anisotropy (FA, the degree of directionality of water diffusion) and mean diffusivity (MD, which describes the overall water diffusion) for 28 WM pathways of interest (POIs). Structural Metrics: Structural MRI data will be used to calculate cortical thickness for 148 regions of interest (ROIs), and to calculate 11 global brain measures (total cortical and subcortical volumes, mean cortical thickness, total cortical surface area, total cortical and cerebellar GM volumes, total cortical and cerebellar WM volumes, brain stem volume, corpus callosum volume, and white matter hypointensities). Brain Age Model: Mean FA/MD values for each POI, cortical thickness values for each ROI, global brain measures, and covariates will be used to predict brain age via relevance vector regression, a machine learning model that applies regularization through a Gaussian prior. We will use 10-fold cross validation to examine model generalizability and mean absolute error to quantify model performance. The trained model with then be applied to the DOD-ADNI dataset of Veterans with and without a history of TBI to determine if there are between-group differences in brain-PAD (calculated as predicted age minus chronological age). Multiple linear regression will be used to determine if brain-PAD is associated with cognitive functioning, psychological functioning, and/or vascular burden. Implications: The ADNI and DOD-ADNI datasets offer the unique opportunity to compare the brains of older adults with and without a history of TBI. The remoteness of the injuries allows us to examine a very important question of how injuries sustained during early adulthood may manifest decades later and interact with brain aging. As our Veterans age, it is important to understand how injuries experienced while serving will influence cognitive aging and therefore their ability to function independently so that we might inform targeted prevention and intervention efforts. |
