Longitudinal tau aggregation, atrophy, and cognitive decline in Alzheimer's disease.

In the absence of disease-modifying therapies for Alzheimer disease, there is a critical need to identify modifiable risk factors that may delay the progression of Alzheimer disease. To examine whether physical activity moderates the association of β-amyloid (Aβ) burden with longitudinal cognitive decline and neurodegeneration in clinically normal individuals and to examine whether these associations are independent of vascular risk. This longitudinal observational study included clinically normal participants from the Harvard Aging Brain Study. Participants were required to have baseline Aβ positron emission tomography data, baseline medical data to quantify vascular risk, and longitudinal neuropsychological and structural magnetic resonance imaging data. Data were collected from April 2010 to June 2018. Data were analyzed from August to December 2018. Baseline physical activity was quantified with a pedometer (mean steps per day). Baseline Aβ burden was measured with carbon 11-labeled Pittsburgh Compound B positron emission tomography. Cognition was measured annually with the Preclinical Alzheimer Cognitive Composite (PACC; median [interquartile range] follow-up, 6.0 [4.3-6.3] years). Neurodegeneration was assessed with longitudinal structural magnetic resonance imaging (2 to 5 scans per participant; median [interquartile range] follow-up, 4.5 [3.0-5.0] years), with a focus on total gray matter volume and regional cortical thickness. Physical activity and Aβ burden were examined as interactive predictors of PACC decline and volume loss in separate linear mixed models, adjusting for age, sex, education, apolipoprotein E ε4 status, and, where appropriate, intracranial volume. Secondary models adjusted for vascular risk and its interaction with Aβ burden. Of the 182 included participants, 103 (56.6%) were female, and the mean (SD) age was 73.4 (6.2) years. In models examining PACC decline and volume loss, there was a significant interaction of physical activity with Aβ burden, such that greater physical activity was associated with slower Aβ-related cognitive decline (β, 0.03; 95% CI, 0.02-0.05; P < .001) and volume loss (β, 482.07; 95% CI, 189.40-774.74; P = .002). Adjusting for vascular risk did not alter these associations. In these models, lower vascular risk was independently associated with slower Aβ-related PACC decline (β, -0.04; 95% CI, -0.06 to -0.02; P < .001) and volume loss (β, -483.41; 95% CI, -855.63 to -111.20; P = .01). Greater physical activity and lower vascular risk independently attenuated the negative association of Aβ burden with cognitive decline and neurodegeneration in asymptomatic individuals. These findings suggest that engaging in physical activity and lowering vascular risk may have additive protective effects on delaying the progression of Alzheimer disease.

The spatial and temporal patterns of cortical mean diffusivity (cMD), as well as its association with Alzheimer's disease (AD) and suspected non-Alzheimer's pathophysiology (SNAP), are not yet fully understood. We compared baseline (n=617) and longitudinal changes (n=421) of cMD, cortical thickness, and gray matter volume and their relations to vascular risk factors, amyloid beta (Aβ), and tau positron emission tomography (PET), and longitudinal cognitive decline in Aβ PET negative and positive older adults. cMD increases were more sensitive to detecting brain structural alterations than cortical thinning and gray matter atrophy. Tau-related cMD increases partially mediated Aβ-related cognitive decline in AD, whereas vascular disease-related increased cMD levels substantially mediated age-related cognitive decline in SNAP. These findings revealed the dynamic changes of microstructural and macrostructural indicators and their associations with AD and SNAP, providing novel insights into understanding upstream and downstream events of cMD in neurodegenerative disease. Cortical mean diffusivity (cMD) was more sensitive to detecting structural changes than macrostructural factors. Tau-related cMD increases partially mediated amyloid beta-related cognitive decline in Alzheimer's disease (AD). White matter hyperintensity-related higher cMD mainly explained the age-related cognitive decline in suspected non-Alzheimer's pathophysiology (SNAP). cMD may assist in tracking earlier neurodegenerative signs in AD and SNAP.

INTRODUCTIONBaseline and longitudinal characteristics of cerebrospinal fluid (CSF) growth‐associated protein 43 (GAP‐43) and plasma neurofilament light (NfL) and how they correlate interactively with neurodegeneration and cognitive decline in Alzheimer's disease (AD) are not fully understood.METHODSWe investigated dynamic changes of CSF GAP‐43 and plasma NfL across different AD stages and their association with longitudinal neurodegeneration and cognitive decline up to 12 years.RESULTSIndividuals with hippocampal atrophy, AD‐signature cortical thinning, or hypometabolism (N+) had faster plasma NfL increase rates than healthy individuals, regardless of amyloid/tau status. In contrast, none of these N+ imaging indicators correlated with more rapid increases in CSF GAP‐43. Furthermore, CSF GAP‐43 and plasma NfL synergistically predicted subsequent gray matter atrophy, cortical thinning, hypometabolism of the middle temporal region, and cognition.DISCUSSIONCSF GAP‐43‐associated presynaptic loss indicates tau‐dependent early neurodegeneration, whereas the axonal degeneration indicated by plasma NfL is a relatively late atrophy/hypometabolism‐associated fluid neurodegeneration biomarker.HighlightsPlasma neurofilament light (NfL) was increased in N+ or cognitively impaired individuals.Increases in tau‐dependent cerebrospinal fluid CSF growth‐associated protein 43 (GAP‐43) before imaging neurodegeneration indicators.CSF GAP‐43 and plasma NfL are synergistically related to longitudinal neurodegeneration.CSF GAP‐43 and plasma NfL are synergistically related to longitudinal cognitive decline.

To examine the patterns of longitudinal tau accumulation and cortical atrophy and their association in subjects with mild cognitive impairment (MCI). We collected 23 participants (60-89years old, 11 males/12 females) with MCI from the Alzheimer's Disease Neuroimaging Initiative database. All participants underwent 18F flortaucipir (FTP) positron emission tomography (PET) and structural magnetic resonance imaging (MRI) scans at the baseline and follow-up visits (12-36months). General linear models with covariates (baseline age, sex) were used to detect brain areas of significant tau accumulation and atrophy over time. Mediation analysis was employed to explore the potential reason for sequential biomarker changes in MCI progression, adjusting for baseline age, sex, and education level. Voxel-wise tau accumulation in MCI subjects was predominantly located in the inferior temporal cortex, middle temporal cortex, parietal cortex, posterior cingulate, precuneus, and temporoparietal regions (P < 0.001), and MRI atrophy included the inferior-middle temporal lobe, parietal lobe, and precuneus (P < 0.001). Longitudinal FTP accumulation was moderately associated with annualized MRI cortical atrophy (r = 0.409, 95% CI: 0.405-0.414, P < 0.01). Regional analyses indicated significant bivariate associations between annualized MRI cortical atrophy and FTP accumulation (baseline FTP cortical uptake and longitudinal FTP change). The results of the mediation analysis showed that the relationship between baseline FTP uptake and longitudinal cortical atrophy was partly mediated by the longitudinal FTP cortical change (indirect effect: 0.0107, P = 0.04). Our findings provide a preliminary description of the patterns of longitudinal FTP accumulation and annualized cortical atrophy in MCI progression, and MCI subjects with high tau binding levels show an increase risk of longitudinal tau accumulation, atrophy, and cognitive decline. Trial registration NCT00106899. Registered 1 April 2005, https://clinicaltrials.gov/ct2/show/study/NCT00106899.

Reduced plasma desmosterol-to-cholesterol ratio and longitudinal cognitive decline in Alzheimer's disease

Psychosis in Alzheimer’s Disease

PurposeBiomarkers used for predicting longitudinal cognitive change in Alzheimer’s disease (AD) continuum are still elusive. Tau pathology, neuroinflammation, and neurodegeneration are the leading candidate predictors. We aimed to determine these three aspects of biomarkers in cerebrospinal fluid (CSF) and plasma to predict longitudinal cognition status using Alzheimer’s Disease Neuroimaging Initiative (ADNI) cohort.Patients and MethodsA total of 430 subjects including, 96 cognitive normal (CN) with amyloid β (Aβ)-negative, 54 CN with Aβ-positive, 195 mild cognitive impairment (MCI) with Aβ-positive, and 85 AD with amyloid-positive (Aβ-positive are identified by CSF Aβ42/Aβ40 < 0.138). Aβ burden was evaluated by CSF and plasma Aβ42/Aβ40 ratio; tau pathology was evaluated by CSF and plasma phosphorylated-tau (p-tau181); microglial activation was measured by CSF soluble TREM2 (sTREM2) and progranulin (PGRN); neurodegeneration was measured by CSF and plasma t-tau and structural magnetic resonance imaging (MRI); cognition was examined annually over the subsequent 8 years using the Alzheimer’s Disease Assessment Scale Cognition 13-item scale (ADAS13) and Mini-Mental State Exam (MMSE). Linear mixed-effects models (LME) were applied to assess the correlation between biomarkers and longitudinal cognition decline, as well as their effect size on the prediction of longitudinal cognitive decline.ResultsBaseline CSF Aβ42/Aβ40 ratio was decreased in MCI and AD compared to CN, while CSF p-tau181 and t-tau increased. Baseline CSF sTREM2 and PGRN did not show any differences in MCI and AD compared to CN. Baseline brain volumes (including the hippocampal, entorhinal, middle temporal lobe, and whole-brain) decreased in MCI and AD groups. For the longitudinal study, there were significant interaction effects of CSF p-tau181 × time, plasma p-tau181 × time, CSF sTREM2 × time, and brain volumes × time, indicating CSF, and plasma p-tau181, CSF sTREM2, and brain volumes could predict longitudinal cognition deterioration rate. CSF sTREM2, CSF, and plasma p-tau181 had similar medium prediction effects, while brain volumes showed stronger effects in predicting cognition decline.ConclusionOur study reported that baseline CSF sTREM2, CSF, and plasma p-tau181, as well as structural MRI, could predict longitudinal cognitive decline in subjects with positive AD pathology. Plasma p-tau181 can be used as a relatively noninvasive reliable biomarker for AD longitudinal cognition decline prediction.

BackgroundPrevious cross‐sectional studies suggest that higher educational attainment (EA) may mitigate cognitive decline in Alzheimer’s disease (AD). This study systematically investigated the association between EA and longitudinal cognitive decline across various domains and explored its connections with β‐amyloid (Aβ) plaque, tau tangle, gray matter volume (GMV) and glucose metabolism in AD.MethodWe analyzed Aβ‐PET (A), tau‐PET (T), and 3D T1‐MRI images (N) from the ADNI cohort to identify 58 A+/T‐, 77 A+/T+, and 84 cognitively unimpaired (CU) participants without evidence of AD pathology and neurodegeneration (A‐/T‐/N‐/CU). All participants had longitudinal cognitive assessments covering the Preclinical Alzheimer Cognitive Composite (PACC), memory, executive function, language and visuospatial function. Additionally, a subset of 67 individuals had FDG‐PET images. Participants were divided into High‐Edu and Low‐Edu subgroups based on the median years of education of the whole cohort (16 years). Baseline and follow‐up cognitive measurements in various domains were compared between High‐Edu and Low‐Edu individuals. Furthermore, the association of EA, GMV, FDG, Aβ plaque, tau tangle, and cognitive changes were investigated.ResultSlower cognitive decline was specifically observed in PACC and language domains of the A+/T+ group among High‐Edu individuals compared to Low‐Edu individuals (Figure 1). The mediation analysis demonstrated that higher EA attenuated PACC decline through greater GMV in the middle temporal gyrus (MTG) and higher glucose metabolism in the left parahippocampal gyrus. Additionally, it mitigated language decline through greater GMV and glucose metabolism in the left inferior temporal gyrus. Furthermore, slower PACC and language decline in High‐Edu individuals was fully mediated through lower tau tangles in the MTG and left superior temporal gyrus (STG), rather than through an Aβ‐associated pathway (Figure 2). A four‐variable mediation analysis further confirmed that tau tangles and GMV in the MTG independently mediated the association between EA and PACC decline (Figure 3).ConclusionThese findings suggest that higher EA has a prominent protective effect in mitigating PACC and language decline among A+/T+ individuals. The protective effects of higher EA are mediated through the maintenance of greater GMV (or FDG) and lower tau tangles, and these pathways independently contribute to the attenuation of PACC and language decline.

BackgroundCerebrospinal fluid (CSF) concentration of soluble TREM2 (sTREM2), a potential biomarker for microglial activation, is associated with attenuated longitudinal neurodegeneration and cognitive decline in Alzheimer’s disease (AD), but data in early disease are lacking. This study’s purpose was to use longitudinal volumetric imaging to assess the association of sTREM2 with age‐ and preclinical AD‐related grey matter (GM) changes.MethodCognitively unimpaired participants (N = 384; amyloid‐positive N = 82) from the Wisconsin Registry for Alzheimer’s Prevention and Wisconsin ADRC clinical core studies with baseline CSF biomarker and subsequent longitudinal T1‐weighted magnetic resonance imaging data were analyzed. CSF sTREM2 and phosphorylated‐tau181/amyloid‐beta1‐42 ratio (pTau/Aβ42) were measured using the NeuroToolKit panel of robust prototype assays (Roche Diagnostics International Ltd, Rotkreuz, Switzerland). T1‐weighted images were longitudinally registered to intra‐subject templates and segmented to create 58 GM regions of interest (ROIs) via the CAT12 longitudinal segmentation pipeline. Linear mixed‐effects models (random participant intercepts and age slopes) testing a three‐way interaction between time‐varying age, sTREM2, and pTau/Aβ42 to predict regional grey matter changes with all potential two‐way interactions and simple effects (adjusted for gender, years of education, intracranial volume, and head coil) were tested. In the event of a non‐significant three‐way interaction, the three‐way interaction was dropped and the model was reinterpreted. Statistical significance was considered at p &lt; .05, uncorrected for multiple comparisons.ResultAge had a negative effect on regional GM volume across the brain and showed widespread interactions with pTau/Aβ42, indicative of accelerated decline with AD pathology. Negative three‐way interactions between sTREM2, pTau/Aβ42, and age were evident in the angular, supramarginal, lingual, and middle occipital gyri, predicting accelerated AD‐related longitudinal neurodegeneration with higher sTREM2 concentration. Negative two‐way interactions between pTau/Aβ42 and sTREM2 were evident in the supplementary motor cortex and superior frontal gyrus, indicating worse effects of AD‐pathology with higher sTREM2, regardless of age.ConclusionOverall, higher sTREM2 may be associated with accelerated AD‐related neurodegeneration over time in the context of preclinical AD, particularly in posterior ROIs. Higher sTREM2 and underlying microglial activation may denote individuals at higher risk of experiencing the deleterious effects of early AD pathology on the brain.

BackgroundActivation of microglial cells in the brain, more commonly known as neuroinflammation, has often been linked to the pathophysiology of Alzheimer’s disease (AD). However, how microglial activation is associated with longitudinal tau tangle accumulation and consequent cognitive decline is poorly understood. Here, we aimed to investigate whether baseline microglial activation impacts tau tangle deposition and cognitive decline in individuals across the AD continuum.MethodWe assessed 92 individuals from the TRIAD cohort (57 cognitively unimpaired and 35 cognitively impaired) with available baseline [11C]PBR28‐PET, a measure of microglial activation and [18F]NAV4694 Aß‐PET, and longitudinal [18F]MK6240 Tau‐PET (mean follow‐up time = 1.93 years) and Mini‐Mental State Exam (MMSE) (mean follow‐up time = 1.84 years). We performed voxel‐wise associations using linear regressions accounting for age and sex and adjusted for multiple comparisons using Random Field Theory (RFT) (p &lt; 0.05). We used the cuneus and superior temporal cortex as a composite ROI for [11C]PBR28‐PET and Aß‐PET since these regions showed a higher association with longitudinal tau accumulation in the temporal meta‐ROI.ResultVoxel‐wise analysis showed that baseline levels of [11C]PBR28‐PET alone are not sufficient to predict longitudinal tau tangle accumulation (Fig. 1a). However, the interaction between [11C]PBR28‐PET levels and Aß burden predicted an increased accumulation of tau tangle, mainly in the cuneus, inferior frontal and lateral occipital regions (Fig. 1b).Individuals with higher baseline [11C]PBR28‐PET and Aß‐PET levels present higher rates of longitudinal tau accumulation in the temporal meta‐ROI (ß = 0.36, t = 3.46, p = 0.0009; Fig. 1c). Similarly, while [11C]PBR28‐PET levels alone did not correlate with longitudinal changes in MMSE score (ß = ‐0.17, t = ‐1.69, p = 0.10), a significant interaction between [11C]PBR28‐PET and Aß‐PET levels on MMSE annual decline was observed (ß = ‐0.24, t = ‐2.25, p = 0.028; Fig. 1d).ConclusionWe found that baseline levels of microglial activation was associated with longitudinal tau tangle accumulation and cognitive decline in individuals across the AD continuum in the presence of Aß burden. Our results indicate that microglial activation might act potentiating the deleterious effects of Aß on forthcoming tau tangle deposition.

Neuroinflammation is believed to be a key process in Alzheimer’s disease (AD) pathogenesis. Recently, the neutrophil-to-lymphocyte (NLR) and lymphocyte-to-monocyte ratios (LMR) have been proposed to be useful peripheral markers of inflammation. However, it is unclear how these inflammatory ratios relate to AD pathology, such as β-amyloid (Aβ) plaques and tau tangles. Using 18F-florbetapir and 18F-flortaucipir positron emission tomography (PET), we sought to determine how the NLR and LMR are associated with AD pathology both cross-sectionally and longitudinally. We further evaluated associations between the NLR and LMR and longitudinal cognitive decline. Using data from the Alzheimer’s Disease Neuroimaging Initiative, we analyzed blood, PET, and cognitive data from 1544 subjects—405 cognitively normal, 838 with mild cognitive impairment (MCI), and 301 with AD. Associations between the NLR and LMR and Aβ and tau on PET were assessed using ordinary least-squares and mixed-effects regression models, while adjusting for age, sex, years of education, and apolipoprotein E ε2 or ε4 carrier status. Associations between the NLR and LMR and cognitive function, as measured by the AD Assessment Scale-Cognitive Subscale, 13-item version, were also assessed. MCI and AD subjects had higher NLR (p = 0.017, p < 0.001, respectively) and lower LMR (p = 0.013, p = 0.023). The NLR, but not the LMR, was significantly associated with Aβ (p = 0.028), suggesting that higher NLR was associated with greater Aβ deposition in the brain. Neither the NLR nor the LMR was associated with tau deposition (p > 0.05). A higher NLR was associated with greater longitudinal cognitive decline (p < 0.001). A higher ratio of peripheral neutrophils to lymphocytes, possibly reflecting an imbalance in innate versus adaptive immunity, is related to greater Aβ deposition and longitudinal cognitive decline. As the field moves toward blood-based biomarkers of AD, the altered balance of innate versus adaptive immunity could be a useful biomarker of underlying pathology and may also serve as a potential therapeutic target.

Alzheimer’s disease (AD) is the most common cause of dementia and threatens the health of millions of people. Early stage diagnosis of AD is critical for improving clinical outcomes and longitudinal magnetic resonance imaging (MRI) data collection can be used to monitor the progress of each patient. However, missing data is a common problem in longitudinal AD studies. The main factors come from subject dropouts and failed scans. This hinders the acquisition of longitudinal sequences that consist of multi-time-point magnetic resonance (MR) images at relatively uniform intervals. In this paper, we present a generative adversarial convolutional network to predict missing structural MRI data. In particular, we include multiple MRI scans as a temporal sequence collected at different times and determine the spatio-temporal relationship between the different scans in the proposed network. We adopt residual bottlenecks in the generator to decrease parameter values and deepen the network. In order to make full use of the longitudinal information, our discriminator classifies not only real MR images from generated MR images, but also fake sequences from real sequences in which the longitudinal MR images for all time points come from the dataset, only the last MR image comes from the generator. Results of our experiment show that our method performs more accurately for the longitudinal structural MRI data prediction of a brain afflicted with AD.

BackgroundThe pathological hallmarks of Alzheimer's disease (AD) include the amyloid-β (Aβ) plaques and phosphorylated tau (p-tau) forming neurofibrillary tangles. Understanding the pathophysiological cascade related to Aβ and tau process is crucial.ObjectiveTo investigate the impact of Aβ positron emission tomography (PET) and cerebrospinal fluid (CSF) p-tau on tau pathology and cognitive decline in AD.MethodsWe analyzed 319 older individuals from the Alzheimer's Disease Neuroimaging Initiative (ADNI) who underwent Aβ (18F-florbetapir or 18F-florbetaben) and tau (18F-flortaucipir) PET scans, along with CSF and cognitive assessments. Aβ positivity (A+) was determined by global standardized uptake value ratio thresholds of ≥1.11 for 18F-florbetapir or ≥1.08 for 18F-florbetaben, while p-tau positivity (T+) was defined as CSF p-tau181 levels ≥23 pg/ml. Linear mixed regression models were used to assess the effects of PET Aβ and CSF p-tau181 levels on tau accumulation in predefined Braak regions and cognitive function over time.ResultsOur results revealed significant differences in PET tau pathology and cognitive decline between A + and A- individuals. We observed that interactions between Aβ and p-tau proteins were associated with tau accumulation and cognitive decline. Additionally, A-/T + individuals exhibited higher levels of tau accumulation in all Braak regions compared to A-/T- counterparts, suggesting a potential independent role of p-tau in tau pathology in the absence of Aβ.ConclusionsOur findings suggest that Aβ positivity and elevated CSF p-tau181 levels were associated with tau accumulation and cognitive decline, highlighting the relevance of soluble p-tau as a potential biomarker for further investigation.

Alzheimer disease (AD) is the main medical problem in adults with Down syndrome (DS). However, the associations of age, intellectual disability (ID), and clinical status with progression and longitudinal cognitive decline have not been established. To examine clinical progression along the AD continuum and its related cognitive decline and to explore the presence of practice effects and floor effects with repeated assessments. This is a single-center cohort study of adults (aged >18 years) with DS with different ID levels and at least 6 months of follow-up between November 2012 and December 2021. The data are from a population-based health plan designed to screen for AD in adults with DS in Catalonia, Spain. Individuals were classified as being asymptomatic, having prodromal AD, or having AD dementia. Neurological and neuropsychological assessments. The main outcome was clinical change along the AD continuum. Cognitive decline was measured by the Cambridge Cognitive Examination for Older Adults With Down Syndrome and the modified Cued Recall Test. A total of 632 adults with DS (mean [SD] age, 42.6 [11.4] years; 292 women [46.2%]) with 2847 evaluations (mean [SD] follow-up, 28.8 [18.7] months) were assessed. At baseline, there were 436 asymptomatic individuals, 69 patients with prodromal AD, and 127 with AD dementia. After 5 years of follow-up, 17.1% (95% CI, 12.5%-21.5%) of asymptomatic individuals progressed to symptomatic AD in an age-dependent manner (0.6% [95% CI, 0%-1.8%] for age <40 years; 21.1% [95% CI, 8.0%-32.5%] for age 40-44 years; 41.4% [95% CI, 23.1%-55.3%] for age 45-49 years; 57.5% [95% CI, 38.2%-70.8%] for age ≥50 years; P < .001), and 94.1% (95% CI, 84.6%-98.0%) of patients with prodromal AD progressed to dementia with no age dependency. Cognitive decline in the older individuals was most common among those who progressed to symptomatic AD and symptomatic individuals themselves. Importantly, individuals with mild and moderate ID had no differences in longitudinal cognitive decline despite having different performance at baseline. This study also found practice and floor effects, which obscured the assessment of longitudinal cognitive decline. This study found an association between the development of symptomatic AD and a high risk of progressive cognitive decline among patients with DS. These results support the need for population health plans to screen for AD-related cognitive decline from the fourth decade of life and provide important longitudinal data to inform clinical trials in adults with DS to prevent AD.

INTRODUCTIONSleep disturbances are common in Alzheimer's disease (AD) and may reflect pathologic changes in brain networks. To date, no studies have examined changes in sleep functional connectivity (FC) in AD or their relationship with network hyperexcitability and cognition.METHODSWe assessed electroencephalogram (EEG) sleep FC in 33 healthy controls, 36 individuals with AD without epilepsy, and 14 individuals with AD and epilepsy.RESULTSAD participants showed increased gamma connectivity in stage 2 sleep (N2), which was associated with longitudinal cognitive decline. Network hyperexcitability in AD was associated with a distinct sleep connectivity signature, characterized by decreased N2 delta connectivity and reversal of several connectivity changes associated with AD. Machine learning algorithms using sleep connectivity features accurately distinguished diagnostic groups and identified “fast cognitive decliners” among study participants who had AD.DISCUSSIONOur findings reveal changes in sleep functional networks associated with cognitive decline in AD and may have implications for disease monitoring and therapeutic development.HighlightsBrain functional connectivity (FC) in Alzheimer's disease is altered during sleep.Sleep FC measures correlate with cognitive decline in AD.Network hyperexcitability in AD has a distinct sleep connectivity signature.