Network homeostasis: functional brain network alterations and relapse in remitted late-life depression

Network homeostasis: functional brain network alterations and relapse in remitted late-life depression | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Network homeostasis: functional brain network alterations and relapse in remitted late-life depression andrew gerlach, Helmet T Karim, antonija kolobaric, brian boyd, and 5 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5005391/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract In this study, we aim to identify neurobiological factors that predict relapse risk in late-life depression (LLD). We recruited 145 older adults (age ≥ 60): 102 recently remitted LLD participants and 43 healthy comparisons. Participants underwent baseline MRI and evaluation of depression symptoms/status for up to 2 years. We evaluated intrinsic network connectivity for 111 participants (39 healthy comparisons, 47 stable remitted, 25 relapsed). LLD participants had lower connectivity primarily within and between the default mode (DMN), somatomotor, and visual networks and higher connectivity between the DMN and executive control network. Lower connectivity of DMN to somatomotor and salience networks was associated with relapse. Notably, the network structure of relapsed participants was more similar to healthy comparisons than stable remitted. These findings indicate that remission is associated with persistent functional network alterations while vulnerability to relapse may be associated with a failure to establish a new stable homeostatic functional network structure. Health sciences/Diseases/Psychiatric disorders/Depression Biological sciences/Neuroscience/Emotion Figures Figure 1 Figure 2 Figure 3 Figure 4 1. Introduction Late-life depression (LLD) is a highly recurrent illness 1 , 2 associated with disability and increased mortality 3 . Even with successful treatment, over half of patients will re-experience depression within four years 4 , mostly within the first two years 5 . Continued exposure to depression increases the risk of metabolic disease, cognitive decline, and death 6 – 8 . We previously conceptualized recurrence of LLD through a model of neural network homeostatic dysregulation 9 (Fig. 1 ). Dynamic neural circuitry copes with stressors through allostatic responses encompassing both behavioral and physiologic adaptations 10 . While remission of depressive symptoms may reestablish homeostasis, remitted individuals may show long-term state-independent, altered brain network configurations 11 . The homeostatic equilibrium of older remitted depressed individuals may be tenuous and further challenged by the degradation of appropriate allostatic responses in older age. This may render older adults particularly vulnerable to recurrence of depression. Our homeostatic disequilibrium hypothesis of LLD recurrence proposes that vulnerability to new depressive episodes stems from chronic fragility in neural network homeostasis persisting in remission. Stressors evoke aberrant neural responses that disrupt homeostasis, alter network function, and subsequently result in maladaptive cognitive and behavioral activity 2 , 9 . This leads to a cycle of further disruption in neural function and, eventually, precipitation of another major depressive episode. This model is consistent with recent dynamical systems views of psychiatric disorders 12 , 13 , which postulates depression and health as attractor states. The attractor landscape defines the likelihood of relapse (conceptualized as critical transition), though the landscape is defined in abstract terms. Depression has been conceptualized as a disruption of large-scale brain networks, particularly the default mode network (DMN), executive control network (ECN), and salience network (SN) 9 , 14 , 15 . These networks are altered in both aging and depression 2 . More recently, the somatomotor network (SMN) and visual network have also been implicated in depression and specifically LLD 16 , 17 . These networks are typically characterized with resting state functional magnetic resonance imaging (fMRI) studies. In addition to evidence for alterations in depression, we and others have demonstrated that, while resting state functional connectivity (FC) is affected by antidepressant treatment 18 , alterations persist in remission compared to healthy comparison older adults 19 . Thus, FC within and between large scale intrinsic brain networks may comprise a meaningful characterization of the neural landscape that define susceptibility or resilience to critical transitions leading to relapse. In this study, we aimed to identify neurobiological factors associated with LLD recurrence risk by conducting a two-year longitudinal study of remitted LLD participants and health comparison older adults. This is one of the first studies to prospectively assess neural markers of relapse and recurrence in depression 20 and, to our knowledge, the first to do so in LLD. For this analysis, we compared resting state FC of seven canonical intrinsic brain networks in older healthy comparisons and participants recently remitted form an acute depressive episode. While our homeostatic disequilibrium model is primarily concerned with dynamic evolution of depressive symptoms and neural architecture, we can use the model to make specific predictions about how stable remission and relapse may differ from each other and healthy comparisons. Specifically, we would expect healthy comparisons to model a stable landscape. Stable remission should present as a similarly stable landscape, while the landscape of eventual relapse should present with greater differences that render it unstable. However, we also expect differences between the remitted landscape, regardless of eventual relapse status, and healthy comparisons. We hypothesized that: 1) within and between network FC, which we define as the neural landscape, will differ between healthy comparisons and remitted participants (regardless of eventual relapse status); 2) remitted participants who relapse will exhibit a different neural landscape compared to participants with stable remission, reflecting unstable vs. stable homeostatic setpoints; 3) the neural landscape in stable remission will be more similar to healthy comparisons than relapse will; and 4) these differences will largely lie in the DMN, ECN, and SN. 2. Results Demographic and clinical characteristics are provided in Table 1 . Three participants did not have MRI data, and 31 participants had excessive motion (greater than 20% of volumes had root mean square motion > 0.5 mm) during the resting state scans, yielding a sample of 111 participants (39 healthy comparison, 47 stable remitted, 25 relapsed) for analysis. The group composition differed by site, with significantly more LLD participants recruited at VUMC due to differing COVID-19 restrictions at Pitt and UIC early in the study. Sex and education differed between groups and was controlled for in all models along with age, race, and site. Table 1 Participant summary. ITP – initial treatment phase; LLD – late-life depression; MADRS – Montgomery Asberg Depression Rating Scale; SD – standard deviation. Overall (n = 145) Healthy comparison (n = 43) LLD (n = 102) Difference, p value Age (mean, SD) 66.9 (4.8) 66.6 (5.4) 67.0 (4.6) \(\:{t}_{68}=-0.35,\:p=0.723\) Sex (female, %) 95 (65.5%) 21 (48.8%) 74 (72.5%) \(\:{\chi\:}_{1}^{2}=5.31,\:p=0.021\) Race 2 White Black Asian Native Am. Multiracial Other (7) 119 (82.1%) 15 (10.3%) 4 (2.8%) 1 (0.7%) 5 (3.4%) 1 (0.7%) 31 (72.1%) 6 (14.0%) 4 (9.3%) 1 (2.3%) 1 (2.3%) 0 (0.0%) 88 (86.3%) 9 (8.8%) 0 (0.0%) 0 (0.0%) 0 (0.0%) 1 (1.0%) \(\:{\chi\:}_{1}^{2}=3.23,\:p=0.072\) Education (mean, SD) 16.1 (2.3) 16.7 (2.0) 15.8 (2.4) \(\:{t}_{94}=2.19,\:p=0.031\) Site VUMC Pitt UIC 77 (53.1%) 34 (23.4%) 34 (23.4%) 14 (32.6%) 14 (32.6%) 15 (34.9%) 63 (61.8%) 20 (19.6%) 19 (18.6%) \(\:{\chi\:}_{2}^{2}=10.43,\:p=0.005\) Baseline MADRS (mean, SD) 4.1 (3.6) 0.9 (1.5) 5.5 (3.3) \(\:{t}_{139}=11.35,\:p<0.001\) Treatment type (ITP, %) - - 58 (57%) - Healthy comparison (n = 43) Stable Remitted (n = 59) Relapsed (n = 43) Difference , p value Age (mean, SD) 66.6 (5.4) 66.9 (4.7) 67.0 (4.4) \(\:{F}_{2}=0.07,\:p=0.930\) Sex a (female, %) 21 (48.8%) 41 (69.5%) 33 (76.7%) \(\:{\chi\:}_{2}^{2}=8.11,\:p=0.017\) Race b White Black Asian Native Am. Multiracial Other (7) 31 (72.1%) 6 (14.0%) 4 (9.3%) 1 (2.3%) 1 (2.3%) 0 (0.0%) 51 (86.4%) 5 (4.9%) 0 (0.0%) 0 (0.0%) 2 (3.4%) 1 (1.7%) 37 (86.0%) 4 (9.3%) 0 (0.0%) 0 (0.0%) 2 (4.7%) 0 (0.0%) \(\:{\chi\:}_{2}^{2}=4.14,\:p=0.126\) Education (mean, SD) 16.7 (2.0) 15.9 (2.1) 15.6 (2.7) \(\:{F}_{2}=2.32,\:p=0.102\) Site VUMC Pitt UIC 14 (32.6%) 14 (32.6%) 15 (34.9%) 31 (52.5%) 14 (23.7%) 14 (23.7%) 32 (74.4%) 6 (14.0%) 5 (11.6%) Baseline MADRS (mean, SD) 0.9 (1.5) 5.4 (3.6) 5.7 (3.0) \(\:{F}_{2}=38.01,\:p<0.001\) Treatment type (ITP, %) - 30 (50.8%) 28 (65.1%) \(\:{\chi\:}_{1}^{2}=4.17,\:p=0.041\) a. Self-reported b. Race coded as binary variable for analysis with groups of black/Asian/Native American/Multiracial/Other and white. 2.1. All remitted LLD participants vs. healthy comparisons Results are summarized in Fig. 2 and full detail is provided in supplemental Table 1. LLD participants exhibited lower FC than healthy comparisons within the DMN ( \(\:HC=7.94,\:p=0.0008\) ), SN ( \(\:HC=10.4,\:p=0.0001\) ), visual ( \(\:HC=7.18,p=0.002\) ), and SMN ( \(\:HC=20.0,\:p<0.0001\) ), as well as between the visual network and DMN ( \(\:HC=9.16,p=0.0002\) ), limbic ( \(\:HC=7.86,p=0.0008\) ), and SMN ( \(\:HC=18.9,p<0.0001\) ). FC between the DMN and ECN was higher in LLD participants than healthy comparisons ( \(\:HC=12.8,p<0.0001\) ). 2.2. Stable remitted vs. relapse vs. healthy comparisons Results are summarized in Fig. 3 and full detail is provided in supplemental Table 2. For the three group comparison, the DMN exhibited significant FC differences between all 7 networks: ECN ( \(\:HC=10.2,p=0.0001\) ), SN ( \(\:HC=9.43,p=0.0001\) ), dorsal attention network (DAN) ( \(\:HC=12.6,p<0.0001\) ), limbic ( \(\:HC=7.93,p=0.0008\) ), visual ( \(\:HC=11.5,p<0.0001\) ), SMN ( \(\:HC=9.24,p=0.0002\) ) and within DMN ( \(\:HC=10.4,p=0.0001\) ). Additional differences were observed between the ECN and visual ( \(\:HC=10.7,p<0.0001\) ) and SMN ( \(\:HC=8.00,p=0.0007\) ), between the DAN and SMN ( \(\:HC=7.23,p=0.0017\) ), and between visual and SMN ( \(\:HC=7.43,p=0.0015\) ). Only these network pairs were tested for pairwise group differences to determine which groups exhibited differences and directionality of effects. 2.2.1. Stable remitted vs. relapse participants Results are summarized in Fig. 4 A and full detail is provided in supplemental Table 3. LLD participants who went on to relapse exhibited lower FC than stable remitted participants between the DMN and SN ( \(\:HC=12.9,p<0.0001\) ), DMN and SMN ( \(\:HC=17.0,p<0.0001\) ), and ECN and SMN ( \(\:HC=7.67,p=0.0014\) ). The DMN-SN ( \(\:HC=13.6,p<0.0001)\) and DMN-SMN ( \(\:HC=19.3,p<0.0001\) ) connectivities were also indicative of time to relapse in the survival analysis. 2.2.2. Stable remitted and relapse vs. healthy comparisons Results are summarized in Fig. 4 B and 4 C and full detail is provided in supplemental Tables 4 and 5. Stable remitted participants exhibited widespread differences from healthy comparisons, including greater FC between DMN and ECN ( \(\:HC=15.7,p<0.0001\) ), SN( \(\:HC=9.18,p=0.0002\) ), and SMN ( \(\:HC=10.3,p=0.0001\) ) and lower FC within the DMN ( \(\:HC=7.21,p=0.0017\) ), between DMN and visual ( \(\:HC=12.2,p<0.0001\) ), and between visual and SMN ( \(\:HC=19.9,p<0.0001\) ). By comparison, participants who go on to relapse have far fewer differences from healthy comparisons: greater FC between DMN and ECN ( \(\:HC=9.64,p=0.0001\) ) and lower FC between SMN and DMN ( \(\:HC=8.87,p=0.0003\) ) and visual ( \(\:HC=17.2,p<0.0001\) ). To test this further, we computed participant-wise similarity matrices for the functional connectomes and compared the similarities between healthy comparison and stable remitted to the similarities between healthy comparison and relapse. Overall, the neural landscape of the relapse group was more similar to healthy comparisons than the neural landscape of the stable remission group was ( \(\:t=-2.51,p=0.0123\) ). 3. Discussion In this study of relapse and recurrence in late-life depression, we found robust differences in network connectivity between relapse, stable remission, and heathy comparisons, differences that centered heavily on the DMN. The remitted LLD group as a whole exhibited lower within and between network connectivity of the DMN, visual, and SMN than healthy comparisons, with the exception of greater connectivity between DMN and ECN. Stable remission and relapse differed primarily by lower connectivity between DMN and SN and SMN in the relapse group. Overall, the connectivity of participants that went on to relapse was more similar to healthy comparisons than was the connectivity of the stable remitted participants. Consistent with our first hypothesis, the remitted LLD group as a whole showed significant differences in network connectivity from healthy comparisons, reflecting a new homeostatic setpoint associated with remission. The DMN featured prominently in these differences, consistent with previous working implicating aberrant DMN connectivity in both adult and geriatric depression 2 , 21 . The finding of greater connectivity between the DMN and ECN in the remitted LLD group partially replicates findings in younger adults that connectivity between specific DMN-ECN regions was elevated in relapsing participants (though not in stable remitted participants) 22 , 23 . There is also evidence that DMN-ECN connectivity is higher in acutely depressed individuals 21 , while our previous review identified a positive association between DMN-ECN connectivity and antidepressant treatment response 24 . The coherence of the DMN and ECN appears to be a key feature of depression, though the precise role remains unclear. The SMN and visual network also showed lower within-network connectivity in LLD, which is consistent with findings from a large Chinese consortium 16 , 17 , though at least one small study has reported the opposite effect in the visual network 25 . Overall, remitted LLD participants show robust differences from healthy comparisons, supporting that stable remission from depression is not simply a “return to normal,” but is associated with a new configuration of the neural landscape. The neural landscape also differed between stabled remitted and relapsed participants. DMN-SN connectivity was lower in relapsed participants than in stable remitted, and was also associated with time to relapse. This is consistent with a recent study in midlife depression that reported relapse was associated with lower connectivity between the DMN and portions of the SN 26 , but stands in contrast to another a recent finding (albeit in only 9 relapsed participants) showing higher FC between the right anterior insula (a SN hub) and the subcallosal cingulate (a DMN hub) predicted depression recurrence 20 . Interestingly, lower DMN-SN connectivity has also been reported in acutely depessed individuals 27 . Lower SMN connectivity to the DMN and ECN was also associated with relapse which is the first finding to our knowledge relating SMN connectivity to relapse and requires further exploration. Contrary to our third hypothesis, participants who would go on to relapse displayed a functional connectivity profile more similar to healthy comparisons than the stable remitted participants did. This finding may indicate that significant reconfigurations of the neural landscape in context of successful antidepressant treatment are required to result in stable remission. There is wide-spread evidence for this reconfiguration across the DMN, ECN, and SN 28 – 32 . This type of dynamic network reconfiguration (as opposed to a return to baseline) is consistent with work in learning reversal, showing that new circuits that will override the learned behavior, rather than a reversal of the learned circuitry 33 . As a corollary, a partial return to baseline (i.e., failure to establish a new stable setpoint) may represent an unstable equilibrium and thus susceptibility to relapse, especially in the presence of other persistent alterations. This may be evidenced in our study by relapsed participants showing greater similarity of functional network organization to healthy comparisons than stable remitted participants. Our final hypothesis—that the key networks of the LLD neural landscape are the DMN, ECN, and SN—appears to be mostly true. The DMN showed widespread and persistent differences between the groups, underscoring its crucial role in the neural underpinnings of late-life depression. Importantly, these differences did not lie just within the DMN, but often in the interaction between the DMN and other networks. The ECN and SN did not differ as robustly as the DMN, though their connectivity with the DMN served as important differentiators between healthy comparisons and LLD, and between stable remission and relapse, respectively. Connectivity of the SMN and visual networks also differed frequently between groups, adding to a growing body of evidence that these unimodal networks are also involved in depression 17 . The prospective design and 4 month inclusion cutoff following remission are significant strengths of our study, providing ecological validity as evidenced by the excellent agreement with previously observed relapse rates of 43% within two years 34 . This study has several limitations. Our ability to assess the temporal stability of neural networks is hampered by analyzing only baseline neuroimaging. We do not have pretreatment imaging data in the LLD group. Sample size is moderate, although above average for neuroimaging of clinical populations and sufficiently powered for our analytic method that leverages omnibus testing to reduce dimensionality. Remitted participants did not receive uniform antidepressant treatment; while a general treatment algorithm was followed, medications were individually tailored. However, this resulted in a high rate of remission (73% across all three sites) and provides better generalization/clinical translation. The delineation between the stable remitted and relapsed groups was subject to right-censoring; while all participants had ≥ 8 months of follow up at the time of analysis, most participants had a longer duration (up to 2 years). We were insufficiently powered to analyze a uniform cutoff of relapse within 8 months, which would result in only 13 relapsed participants. Since relapse tends to occur sooner rather than later (e.g., only 29% of relapsed participants with two years of data relapsed after one year) and 85% of the participants had ≥ 1 year of follow up, we believe our inclusive approach represents a better approximation to the “true” relapse group that we would observe with 2 years of longitudinal data for all participants. In this study we chose to define the neural landscape in terms of within and between network connectivity using the canonical seven Yeo networks. While this choice is well-justified, there are models/evidence that support other meaningful organizational levels of inquiry (e.g., subnetworks or specific regions, especially subcortical regions). Restricting our analysis to 7 networks allowed for cortex-wide coverage while also minimizing the number of comparisons. This investigation is focused on resting state fMRI; future investigations will tests the effect of task-based fMRI or structural connectivity (e.g., white matter hyperintensities, diffusion imagining measures). In conclusion, we identified robust differences in the functional connectome between healthy comparison and remitted participants in late-life, as well as differences between participants who relapsed and those who remained depression-free. These differences were apparent at the network level and were robust to parcellation scheme. These findings, consistent with our proposed disruption of neural homeostasis model of recurrence and relapse in late-depression, may be used to identify depressed older adults at higher risk of relapse and to adequately tailor preventative interventions to prevent the burden of additional depressive episodes. 4. Methods 4.1. Overview Participants were enrolled in the multisite REMBRANDT Study (Recurrence markers, cognitive burden, and neurobiological homeostasis in late-life depression) at Vanderbilt University Medical Center (VUMC), University of Pittsburgh (Pitt.), and University of Illinois – Chicago (UIC). A full description of the study design and rationale has been published previously 35 . Briefly, LLD participants enter through an Initial Treatment Phase (ITP) if currently depressed, or are recruited directly into the longitudinal phase if they recently (< 4 months) remitted to clinical treatment. LLD participants were treated for up to 20 weeks with an algorithm informed by STAR*D 36 and the Duke Neurocognitive Outcomes of Depression in the Elderly studies 37 . Individuals who did not remit were referred for clinical care. Remission was defined as Montgomery-Åsberg Depression Rating Scale 38 (MADRS) score of ≤ 10 concluding the ITP and longitudinal baseline visit, which must occur within 4 months (but no sooner than 1 month) following remission. Remitted LLD and healthy comparison participants entered the two-year longitudinal phase involving scheduled contact every 2 months. This study was approved by the institutional review boards at all three sites. All participants provided informed consent. A CONSORT diagram for the study is shown in Supplemental Fig. 1. 4.2. Participants We enrolled 145 participants, including 102 remitted LLD participants and 43 healthy comparison participants who completed baseline neuroimaging and at least 8 months of clinical follow-up. Inclusion criteria for all participants were: age ≥ 60, fluent in English, Montreal Cognitive Assessment (MoCA) ≥ 24 or MoCA-BLIND ≥ 18. Further inclusion criteria for LLD: current diagnosis of recurrent major depressive disorder (MDD) based on the Diagnostic and Statistical Manual of Mental Disorders (DSM-5), and MADRS ≥ 15 for depressed participants entering the ITP; remission from current MDD episode and MADRS ≤ 10 for LLD participants entering the longitudinal phase. Exclusion criteria for all participants: current or past Axis I diagnosis except generalized anxiety disorder, panic disorder, or simple phobias; history of substance dependence/abuse in the past year; acute grief or suicidality; current or past psychosis; neurological disorders, including dementia; unstable medical illness requiring treatment; MRI contraindications, electroconvulsive therapy in the past 6 months; current brain stimulation or ketamine/esketamine treatment. Healthy comparison participants were excluded for MADRS scores > 8, current or past depression diagnosis or use of psychotropic medication for psychiatric symptoms. Past brief therapy for specific challenges or losses was allowable. We collected demographic, clinical, neuropsychological and MRI data from all participants. While data collection is currently ongoing, we have at least 8 months of longitudinal clinical data on all participants. Based on the available longitudinal data, we further stratified the LLD group into 59 stable remitted participants and 43 relapsed participants (see Assessments for definition). 4.3. Assessments Depression severity was assessed at baseline and every two months throughout the study. Remitted LLD participants who maintained MADRS ≤ 15 over the duration of the study were classified as stable remitted . Initially remitted participants who experienced a MADRS > 15 plus DSM-5 MDD criteria for at least two weeks were classified as relapsed. In this study, we do not differentiate between relapse (reoccurrence of symptoms from previous episode) and recurrence (occurrence of a new depressive episode) since this delineation is often unclear or relies on arbitray time cutoffs. 4.4. MRI Acquisition MRI were acquired 1–4 months after remission from an acute depressive episode at 3T on Philips Elition (VUMC), Siemens Prisma (Pitt) and GE Discovery MR750 System (UIC) scanners using 32-channel head coils following the Adolescent Brain Cognitive Development (ABCD) Study MRI Protocol 39 to harmonize across sites. We collected high resolution structural images with a sagittal T1-weighted Magnetization Prepared Rapid Gradient Echo (MPRAGE) sequence, an axial T2-weighted Fluid Attenuated Inversion Recovery (FLAIR) sequence, resting state with an axial T2* blood oxygen level dependent (BOLD) gradient-echo echoplanar imaging sequence, and a reverse encoded top-up image for correcting susceptibility artifacts matched to the resting sate acquisition parameters. Specific acquisition parameters are provided in Supplemental Tables 4 and 5 of the study protocol 35 . Two separate 5 minute resting state scans (188 volumes per scan) were acquired back-to-back, which has been shown to improve reliability of measures 40 . We assessed scanner variability across the sites using structural 41 and functional phantoms 42 . All sites showed acceptable levels of signal-to-noise, though there were differences by site; therefore, we control for site in all analyses. 4.5. MRI Processing MRI were processed at VUMC using Docker containers in XNAT to ensure reproducibility with in-house scripts utilizing the Statistical Parametric Mapping (SPM12) toolbox 43 except where noted. The FLAIR image was coregistered to the MPRAGE and both images were used in multispectral segmentation into six canonical tissue types, which produces a deformation field for normalizing images into Montreal Neurological Institute (MNI) space. Two were Gaussians used for white matter to account for white matter hyperintensities common in older adults 44 . The resulting grey matter, white matter, and cerebrospinal fluid (CSF) tissue maps were thresholded at 0.1, filled, and smoothed to create an intracranial volume mask used to skull-strip the structural images. Resting state images were slice-time corrected, corrected for susceptibility induced distortions using topup in FSL 45 , realigned to the mean volume using rigid body transformations, skull-stripped using the brain extraction tool in FSL 46 , coregistered to the MPRAGE, normalized to MNI space (2mm 3 isotropic resolution) using the structural deformation field, and smoothed with a Gaussian kernel with 8mm full width at half maximum. The first five principal components of the white matter and CSF, six motion parameters, and sinusoids for bandpass filtering in the 0.008–0.15 Hz frequency range were regressed out of the image. 4.6. Connectome Calculation We calculated FC matrices by extracting the first principal component of processed image time series in predefined parcellation regions and quantifying region-to-region FC as Pearson correlation of the time series. To ensure results are robust to parcellation scheme, we employed three different functionally-defined parcellations: Shen268 47 , Schaefer200, and Schaefer400 48 , resulting in three FC matrices of size 268x268, 200x200, and 400x400 per scan, with each entry referred to as an edge in the connectome. We overlayed the Yeo 7 Network definitions 49 (DMN, ECN, SN, DAN, limbic network, visual network, and SMN) on these atlases and assigned each node to a canonical intrinsic network with a winner-take-all approach. We restrict our network analysis to within and between network connectivity of these 7 networks for a total of 28 “features” of the neural landscape. 4.7. Statistical Analysis We assessed for differences between the healthy comparison and LLD participants (2 group comparison) and between the healthy comparison, stable remitted, and relapsed groups (3 group comparison) using the generalized linear model. Post-hoc tests for significant 3 group comparison results are performed to determine which groups show differences and the directionality of the effect. We also tested time-to-relapse in the LLD group using a Cox proportional hazards model. Age, sex, race, education, and site were controlled for in all models. For this analysis, we employ mass univariate testing at the edge-level (each entry of the FC matrices) and perform inference at higher organizational levels (e.g., network) on subsets of these tests using higher criticism (HC). HC is omnibus test optimal for detecting rare and weak signals 50 , 51 . This is accomplished by comparing the observed p -values to the theoretical null using a modified Kolmogrov-Smirnov statistic; hence HC only tests for difference among a collection of edges — it does NOT offer inference at the edge-level. This approach optimally balances power and specificity for connectome analysis and has recently been applied to neuroimaging 52 – 54 .At the network level, HC is applied only to entries in the FC matrix corresponding to the specified network pairs (i.e., partitioned submatrices for between network connectivity and triangular blocks on the diagonal for within network connectivity). Thus, while we preform primary statistical tests on full connectome matrices with 200 to 400 nodes, we only perform inference on the 7-network connectome with the many region-to-region tests essentially functioning as multiple observations within each network pair. We apply the Benjamini-Hochberg procedure for controlling the False Discovery Rate at \(\:\alpha\:=0.05\) with \(\:m=28\) . We calculate p values for the HC statistic by comparison to a Monte Carlo simulation of 10,000 null distributions for the given number of tests (edges or entries in the FC submatrix). To compare the connectivity profiles of stable remitted and relapse participants to healthy comparisons, we computed a participant similarity matrix. Similarity was calculated as the inverse of the Euclidean distance between the FC matrix entries of two participants (inverse of the Frobenius norm of the difference FC matrix). Similarity between healthy comparisons and stable remitted was compared to similarity between healthy comparisons and relapse with t -tests. Note that age, sex, race, and education were not controlled for in this analysis since each similarity metric compares two participants. We only report network results that were significant across at least two of the parcellations and note results that were significant across all three parcellations. For brevity, we only report median test statistics across the three parcellations in the text with results for all parcellations contained in the supplement. Declarations Disclosures: Olusola Ajilore is a co-founder of Keywise AI, has served as a consultant for Sage Therapeutics and Otsuka, has received honoraria from Boehringer Ingelheim, and is on the advisory board for Blueprint Health and Embodied Labs. Antonija Kolobaric serves as a consultant for Radicle Science. No other authors have disclosures to report. Acknowledgements: We’d like to thank the numerous staff at VUMC, Pitt, and UIC who helped administer the study protocol and process the data. Previous presentation: This work was presented as a poster at conferences for the American Association for Geriatric Psychiatry (March 16, 2024, Atlanta, GA) and the Society for Biological Psychiatry (May 10, 2024, Austin, TX). Data Availability: Deidentified data is available upon reasonable request. The analysis script is available at https://github.com/REMBRANDT-study/resting_state_network_analysis_baseline145. Funding: This work was supported by National of Institute of Health grants R01 MH121619, R01 MH121620, R01 MH121384, K01 MH133913, R01 MH108509, K01 MH122741, T32 MH019986, and the National Center for Advancing Translational Sciences grants UL1 TR000445 and UL1 TR002243. References Taylor Warren D. Depression in the Elderly. N Engl J Med . 2014;371(13):1228–1236. doi: 10.1056/NEJMcp1402180 Szymkowicz SM, Gerlach AR, Homiack D, Taylor WD. Biological factors influencing depression in later life: role of aging processes and treatment implications. Transl Psychiatry . 2023;13(1):1–16. doi: 10.1038/s41398-023-02464-9 Penninx BWJH, Geerlings SW, Deeg DJH, van Eijk JTM, van Tilburg W, Beekman ATF. Minor and Major Depression and the Risk of Death in Older Persons. Arch Gen Psychiatry . 1999;56(10):889–895. doi: 10.1001/archpsyc.56.10.889 Beekman ATF, Geerlings SW, Deeg DJH, et al. The Natural History of Late-Life Depression: A 6-Year Prospective Study in the Community. Arch Gen Psychiatry . 2002;59(7):605–611. doi: 10.1001/archpsyc.59.7.605 Reynolds CF, Dew MA, Pollock BG, et al. Maintenance treatment of major depression in old age. N Engl J Med . 2006;354(11):1130–1138. doi: 10.1056/NEJMoa052619 Penninx BWJH. Depression and cardiovascular disease: Epidemiological evidence on their linking mechanisms. Neurosci Biobehav Rev . 2017;74:277–286. doi: 10.1016/j.neubiorev.2016.07.003 Pan A, Sun Q, Okereke OI, Rexrode KM, Hu FB. Depression and Risk of Stroke Morbidity and Mortality: A Meta-analysis and Systematic Review. JAMA . 2011;306(11):1241–1249. doi: 10.1001/jama.2011.1282 Ganguli M, Du Y, Dodge HH, Ratcliff GG, Chang CCH. Depressive Symptoms and Cognitive Decline in Late Life: A Prospective Epidemiological Study. Arch Gen Psychiatry . 2006;63(2):153–160. doi: 10.1001/archpsyc.63.2.153 Andreescu C, Ajilore O, Aizenstein HJ, et al. Disruption of Neural Homeostasis as a Model of Relapse and Recurrence in Late-Life Depression. Am J Geriatr Psychiatry . 2019;27(12):1316–1330. doi: 10.1016/j.jagp.2019.07.016 McEwen BS, Gianaros PJ. Stress- and Allostasis-Induced Brain Plasticity. Annu Rev Med . 2011;62(1):431–445. doi: 10.1146/annurev-med-052209-100430 Ming Q, Zhong X, Zhang X, et al. State-Independent and Dependent Neural Responses to Psychosocial Stress in Current and Remitted Depression. Am J Psychiatry . 2017;174(10):971–979. doi: 10.1176/appi.ajp.2017.16080974 Scheffer M, Bockting CL, Borsboom D, et al. A Dynamical Systems View of Psychiatric Disorders—Theory: A Review. JAMA Psychiatry . Published online April 3, 2024. doi: 10.1001/jamapsychiatry.2024.0215 Scheffer M, Bockting CL, Borsboom D, et al. A Dynamical Systems View of Psychiatric Disorders—Practical Implications: A Review. JAMA Psychiatry . Published online April 3, 2024. doi: 10.1001/jamapsychiatry.2024.0228 Menon V. Large-scale brain networks and psychopathology: a unifying triple network model. Trends Cogn Sci . 2011;15(10):483–506. doi: 10.1016/j.tics.2011.08.003 Gunning FM, Oberlin LE, Schier M, Victoria LW. Brain-based mechanisms of late-life depression: Implications for novel interventions. Semin Cell Dev Biol . 2021;116:169–179. doi: 10.1016/j.semcdb.2021.05.002 Tan W, Ouyang X, Huang D, et al. Disrupted intrinsic functional brain network in patients with late-life depression: Evidence from a multi-site dataset. J Affect Disord . 2023;323:631–639. doi: 10.1016/j.jad.2022.12.019 Yang H, Chen X, Chen ZB, et al. Disrupted intrinsic functional brain topology in patients with major depressive disorder. Mol Psychiatry . 2021;26(12):7363–7371. doi: 10.1038/s41380-021-01247-2 Karim HT, Andreescu C, Tudorascu D, et al. Intrinsic functional connectivity in late-life depression: trajectories over the course of pharmacotherapy in remitters and non-remitters. Mol Psychiatry . 2017;22(3):450–457. doi: 10.1038/mp.2016.55 Wu M, Andreescu C, Butters MA, Tamburo R, Reynolds CF, Aizenstein H. Default-mode network connectivity and white matter burden in late-life depression. Psychiatry Res Neuroimaging . 2011;194(1):39–46. doi: 10.1016/j.pscychresns.2011.04.003 Dunlop BW, Cha J, Choi KS, Nemeroff CB, Craighead WE, Mayberg HS. Functional connectivity of salience and affective networks among remitted depressed patients predicts episode recurrence. Neuropsychopharmacology . 2023;48(13):1901–1909. doi: 10.1038/s41386-023-01653-w Kaiser RH, Andrews-Hanna JR, Wager TD, Pizzagalli DA. Large-Scale Network Dysfunction in Major Depressive Disorder: A Meta-analysis of Resting-State Functional Connectivity. JAMA Psychiatry . 2015;72(6):603–611. doi: 10.1001/jamapsychiatry.2015.0071 Langenecker SA, Jenkins LM, Stange JP, et al. Cognitive control neuroimaging measures differentiate between those with and without future recurrence of depression. NeuroImage Clin . 2018;20:1001–1009. doi: 10.1016/j.nicl.2018.10.004 Berwian IM, Wenzel JG, Kuehn L, et al. The relationship between resting-state functional connectivity, antidepressant discontinuation and depression relapse. Sci Rep . 2020;10:22346. doi: 10.1038/s41598-020-79170-9 Gerlach AR, Karim HT, Peciña M, et al. MRI predictors of pharmacotherapy response in major depressive disorder. NeuroImage Clin . 2022;36:103157. doi: 10.1016/j.nicl.2022.103157 Eyre HA, Yang H, Leaver AM, et al. Altered resting-state functional connectivity in late-life depression: a cross-sectional study. J Affect Disord . 2016;189:126–133. doi: 10.1016/j.jad.2015.09.011 van Kleef RS, Kaushik P, Besten M, et al. Understanding and predicting future relapse in depression from resting state functional connectivity and self-referential processing. J Psychiatr Res . 2023;165:305–314. doi: 10.1016/j.jpsychires.2023.07.034 Li G, Liu Y, Zheng Y, et al. Large-scale dynamic causal modeling of major depressive disorder based on resting‐state functional magnetic resonance imaging. Hum Brain Mapp . 2019;41(4):865–881. doi: 10.1002/hbm.24845 Vega JN, Taylor WD, Gandelman JA, et al. Persistent Intrinsic Functional Network Connectivity Alterations in Middle-Aged and Older Women With Remitted Depression. Front Psychiatry . 2020;11:62. doi: 10.3389/fpsyt.2020.00062 Cui J, Wang Y, Liu R, et al. Effects of escitalopram therapy on resting-state functional connectivity of subsystems of the default mode network in unmedicated patients with major depressive disorder. Transl Psychiatry . 2021;11(1):634. doi: 10.1038/s41398-021-01754-4 Jiao K, Xu H, Teng C, et al. Connectivity patterns of cognitive control network in first episode medication-naive depression and remitted depression. Behav Brain Res . 2020;379:112381. doi: 10.1016/j.bbr.2019.112381 Workman CI, Lythe KE, McKie S, et al. A novel resting-state functional magnetic resonance imaging signature of resilience to recurrent depression. Psychol Med . 2017;47(4):597–607. doi: 10.1017/S0033291716002567 Liu J, Fan Y, Ling-Li Zeng null, et al. The neuroprogressive nature of major depressive disorder: evidence from an intrinsic connectome analysis. Transl Psychiatry . 2021;11(1):102. doi: 10.1038/s41398-021-01227-8 Barron HC. Neural inhibition for continual learning and memory. Curr Opin Neurobiol . 2021;67:85–94. doi: 10.1016/j.conb.2020.09.007 Deng Y, McQuoid DR, Potter GG, et al. Predictors of recurrence in remitted late-life depression. Depress Anxiety . 2018;35(7):658–667. doi: 10.1002/da.22772 Taylor WD, Ajilore O, Karim HT, et al. Assessing depression recurrence, cognitive burden, and neurobiological homeostasis in late life: Design and rationale of the REMBRANDT study. J Mood Anxiety Disord . Published online November 4, 2023:100038. doi: 10.1016/j.xjmad.2023.100038 Warden D, Rush AJ, Trivedi MH, Fava M, Wisniewski SR. The STAR*D Project results: a comprehensive review of findings. Curr Psychiatry Rep . 2007;9(6):449–459. doi: 10.1007/s11920-007-0061-3 Steffens DC, McQuoid DR, Krishnan KRR. The Duke Somatic Treatment Algorithm for Geriatric Depression (STAGED) approach. Psychopharmacol Bull . 2002;36(2):58–68. Montgomery SA, Asberg M. A new depression scale designed to be sensitive to change. Br J Psychiatry J Ment Sci . 1979;134:382–389. doi: 10.1192/bjp.134.4.382 Casey BJ, Cannonier T, Conley MI, et al. The Adolescent Brain Cognitive Development (ABCD) study: Imaging acquisition across 21 sites. Adolesc Brain Cogn Dev ABCD Consort Ration Aims Assess Strategy . 2018;32:43–54. doi: 10.1016/j.dcn.2018.03.001 Cho JW, Korchmaros A, Vogelstein JT, Milham MP, Xu T. Impact of concatenating fMRI data on reliability for functional connectomics. NeuroImage . 2021;226:117549. doi: 10.1016/j.neuroimage.2020.117549 Gunter JL, Bernstein MA, Borowski BJ, et al. Measurement of MRI scanner performance with the ADNI phantom. Med Phys . 2009;36(6Part1):2193–2205. doi: 10.1118/1.3116776 Friedman L, Glover GH. Report on a multicenter fMRI quality assurance protocol. J Magn Reson Imaging . 2006;23(6):827–839. doi: 10.1002/jmri.20583 Penny WD, Friston KJ, Ashburner JT, Kiebel SJ, Nichols TE. Statistical Parametric Mapping: The Analysis of Functional Brain Images . Elsevier; 2011. Karim HT, Andreescu C, MacCloud RL, et al. The effects of white matter disease on the accuracy of automated segmentation. Psychiatry Res Neuroimaging . 2016;253:7–14. doi: 10.1016/j.pscychresns.2016.05.003 Andersson JLR, Skare S, Ashburner J. How to correct susceptibility distortions in spin-echo echo-planar images: application to diffusion tensor imaging. NeuroImage . 2003;20(2):870–888. doi: 10.1016/S1053-8119(03)00336-7 Smith SM. Fast robust automated brain extraction. Hum Brain Mapp . 2002;17(3):143–155. doi: 10.1002/hbm.10062 Shen X, Tokoglu F, Papademetris X, Constable RT. Groupwise whole-brain parcellation from resting-state fMRI data for network node identification. NeuroImage . 2013;82:403–415. doi: 10.1016/j.neuroimage.2013.05.081 Schaefer A, Kong R, Gordon EM, et al. Local-Global Parcellation of the Human Cerebral Cortex from Intrinsic Functional Connectivity MRI. Cereb Cortex N Y N 1991 . 2018;28(9):3095–3114. doi: 10.1093/cercor/bhx179 Thomas Yeo BT, Krienen FM, Sepulcre J, et al. The organization of the human cerebral cortex estimated by intrinsic functional connectivity. J Neurophysiol . 2011;106(3):1125–1165. doi: 10.1152/jn.00338.2011 Donoho D, Jin J. Higher criticism for detecting sparse heterogeneous mixtures. Ann Stat . 2004;32(3):962–994. doi: 10.1214/009053604000000265 Donoho D, Jin J. Higher Criticism for Large-Scale Inference, Especially for Rare and Weak Effects. Stat Sci . 2015;30(1):1–25. doi: 10.1214/14-STS506 Gerlach AR, Karim HT, Kazan J, Aizenstein HJ, Krafty RT, Andreescu C. Networks of worry-towards a connectivity-based signature of late-life worry using higher criticism. Transl Psychiatry . 2021;11(1):550. doi: 10.1038/s41398-021-01648-5 Wilson JD, Gerlach AR, Karim HT, Aizenstein HJ, Andreescu C. Sex matters: acute functional connectivity changes as markers of remission in late-life depression differ by sex. Mol Psychiatry . Published online July 6, 2023. doi: 10.1038/s41380-023-02158-0 Sundermann B, Feldmann R, Mathys C, et al. Functional connectivity of cognition-related brain networks in adults with fetal alcohol syndrome. BMC Med . 2023;21(1):496. doi: 10.1186/s12916-023-03208-8 Additional Declarations Yes there is potential Competing Interest. Olusola Ajilore is a co-founder of Keywise AI, has served as a consultant for Sage Therapeutics and Otsuka, has received honoraria from Boehringer Ingelheim, and is on the advisory board for Blueprint Health and Embodied Labs. Antonija Kolobaric serves as a consultant for Radicle Science. No other authors have disclosures to report. Supplementary Files supplement.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5005391","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":367671113,"identity":"d1b5a2dc-0d92-458f-ba9d-b64b563a46fa","order_by":0,"name":"andrew gerlach","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA1UlEQVRIiWNgGAWjYHACxgMMDEDE3gDlHyBCD0QLzwEEl0gtEglEapF3731w4EfNHXnzmW/MHt2oYZDju5GAX4vhmeMGB3uOPTOcczvH3DjnGIOxJEEtM9IYDvA2HGacIZ1jJp3bwJC4gaCW+c8YDv5tOGw/Q/IMWEs9QS3yEmwMh4G2JM6Q4AFrSTAgpMWAJ43hsMyxw8kzeNLKgX6RMJx55gEBW9qPMT58U3PYdgb74W2Pc2ps5PmOE7LlAILNBsQS+JWDbWlA1TIKRsEoGAWjABMAAE8hSmr0AysLAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0002-4022-1356","institution":"University of Pittsburgh","correspondingAuthor":true,"prefix":"","firstName":"andrew","middleName":"","lastName":"gerlach","suffix":""},{"id":367671114,"identity":"201f8e72-b5f7-4d83-ab5d-6b3028bc0eec","order_by":1,"name":"Helmet T Karim","email":"","orcid":"https://orcid.org/0000-0002-9286-0694","institution":"University of Pittsburgh","correspondingAuthor":false,"prefix":"","firstName":"Helmet","middleName":"T","lastName":"Karim","suffix":""},{"id":367671115,"identity":"8e536ee6-6791-476f-ae17-c75f7d423c4c","order_by":2,"name":"antonija kolobaric","email":"","orcid":"","institution":"University of Pittsburgh","correspondingAuthor":false,"prefix":"","firstName":"antonija","middleName":"","lastName":"kolobaric","suffix":""},{"id":367671116,"identity":"d2463b51-ea36-4e53-98dd-3e51ff7363e1","order_by":3,"name":"brian boyd","email":"","orcid":"","institution":"Vanderbilt University Medical Center","correspondingAuthor":false,"prefix":"","firstName":"brian","middleName":"","lastName":"boyd","suffix":""},{"id":367671117,"identity":"0883a1bb-28a6-425f-a6f2-6ac5ca781cd4","order_by":4,"name":"Kevin Kahru","email":"","orcid":"","institution":"University of Pittsburgh","correspondingAuthor":false,"prefix":"","firstName":"Kevin","middleName":"","lastName":"Kahru","suffix":""},{"id":367671118,"identity":"2e71f387-0f8e-4a76-9e54-41caf575de72","order_by":5,"name":"Robert Krafty","email":"","orcid":"","institution":"Emory University","correspondingAuthor":false,"prefix":"","firstName":"Robert","middleName":"","lastName":"Krafty","suffix":""},{"id":367671119,"identity":"15d7a7ef-064d-445b-9127-e8e3b2a97feb","order_by":6,"name":"Olusola Ajilore","email":"","orcid":"","institution":"University of Illinois at Chicago","correspondingAuthor":false,"prefix":"","firstName":"Olusola","middleName":"","lastName":"Ajilore","suffix":""},{"id":367671120,"identity":"8b6f87e5-c1f0-48e2-bc54-41f0d63a1bb8","order_by":7,"name":"Warren Talyor","email":"","orcid":"","institution":"Vanderbilt University Medical Center","correspondingAuthor":false,"prefix":"","firstName":"Warren","middleName":"","lastName":"Talyor","suffix":""},{"id":367671121,"identity":"b91ef335-1e2b-43cf-9c37-b716033cb111","order_by":8,"name":"Carmen Andreescu","email":"","orcid":"","institution":"University of Pittsburgh","correspondingAuthor":false,"prefix":"","firstName":"Carmen","middleName":"","lastName":"Andreescu","suffix":""}],"badges":[],"createdAt":"2024-08-30 17:25:11","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5005391/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5005391/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":67117011,"identity":"b768323a-baf6-4b48-b824-2578d5b5a603","added_by":"auto","created_at":"2024-10-21 10:40:31","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":197780,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eConceptualization of recurrence and relapse in late-life depression.\u003c/em\u003e The panel on the left depicts our conceptualization of a simplified healthy “neural landscape” represented by the surface (A), state represented by the ball, and a hypothetical demarcation between healthy and depressed states represented by the dashed line. In the healthy landscape, acute stressors can create a temporary behavioral state characterized by depressive-like symptoms (B), but removal of those stressors allows a return to normal. In LLD (right panel), this landscape become deformed by a range of intrinsic and external factors such that acute stressors are no longer required to maintain a depressive state (C). Successful treatment may shift the state from depressed to healthy. However, this landscape may still differ from the original healthy form and can be stable (e.g., stable remission, D) or unstable (e.g., relapse/recurrence, E).\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-5005391/v1/63f0b0d5f73585d3e7475808.png"},{"id":67117009,"identity":"89ed7f1a-c19e-4b0e-b0ac-ece615bba4e5","added_by":"auto","created_at":"2024-10-21 10:40:31","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":50254,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eNetwork connectivity differences for all remitted late-life depression (LLD) vs. healthy comparisons (Health.)\u003c/em\u003e. \u0026nbsp;Network connectivities greater in the LLD group are shown in red and greater in the healthy comparison group are shown in blue. Color intensity in represents the number of atlases showing the effect with significant results in 3 atlases considered strong and 2 atlases considered moderate (mod.). DAN – dorsal attention network; DMN – default mode network; ECN – executive control network; SMN – somatomotor network; SN – salience network.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-5005391/v1/828b2c0ed9f4ad7a92464831.png"},{"id":67117919,"identity":"7f38504f-f109-42c7-ada2-b1ae6683d13e","added_by":"auto","created_at":"2024-10-21 10:48:31","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":58779,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eNetwork connectivity differences for stable remitted (Rem.) vs. relapse (Rel.) vs. healthy comparisons (Health.)\u003c/em\u003e. There are FC differences between the healthy comparison (Health.), stable remitted (Rem.), and relapsed groups within the DMN, between the DMN and ECN, between the DMN and SN, and between the ECN and SN. Note that these results do not contain information about directionally of effect (i.e., lower or greater FC), which is addressed in post-hoc pairwise tests (see Figure 4). Color intensity in network results and brain maps represents the number of atlases showing the effect with significant results in 3 atlases considered strong and 2 atlases considered moderate (mod.). DAN – dorsal attention network; DMN – default mode network; ECN – executive control network; SMN – somatomotor network; SN – salience network.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-5005391/v1/5b1995f05fee414df11290ba.png"},{"id":67117012,"identity":"3024a040-d07e-485b-a412-777d653e7731","added_by":"auto","created_at":"2024-10-21 10:40:31","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":155940,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003ePairwise network connectivity differences between stable remitted (Rem.), relapse (Rel.), and healthy comparisons (Health.)\u003c/em\u003e.\u003cem\u003e \u003c/em\u003eStable remitted vs. relapse in panel A, healthy comparison vs. stable remitted in panel B, and healthy comparison vs. relapse in panel C. Color intensity represents the number of atlases showing the effect with significant results in 3 atlases considered strong and 2 atlases considered moderate (mod.). DAN – dorsal attention network; DMN – default mode network; ECN – executive control network; SMN – somatomotor network; SN – salience network.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-5005391/v1/92a68ed759b26da0546b8629.png"},{"id":70931961,"identity":"c91fe05d-06c1-464a-8421-e1e2ab003afe","added_by":"auto","created_at":"2024-12-09 10:13:46","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1174645,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5005391/v1/7edd412c-4e40-4d89-ad05-a64e8bdb8a83.pdf"},{"id":67117013,"identity":"55cbf101-a933-4c8a-bdf6-d8bbc8f73337","added_by":"auto","created_at":"2024-10-21 10:40:31","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":237194,"visible":true,"origin":"","legend":"","description":"","filename":"supplement.docx","url":"https://assets-eu.researchsquare.com/files/rs-5005391/v1/dbb10d72242fc130405ae0a4.docx"}],"financialInterests":"\u003cb\u003eYes\u003c/b\u003e there is potential Competing Interest.\nOlusola Ajilore is a co-founder of Keywise AI, has served as a consultant for Sage Therapeutics and Otsuka, has received honoraria from Boehringer Ingelheim, and is on the advisory board for Blueprint Health and Embodied Labs. Antonija Kolobaric serves as a consultant for Radicle Science. No other authors have disclosures to report.","formattedTitle":"Network homeostasis: functional brain network alterations and relapse in remitted late-life depression","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eLate-life depression (LLD) is a highly recurrent illness\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e,\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e associated with disability and increased mortality\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. Even with successful treatment, over half of patients will re-experience depression within four years\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e, mostly within the first two years\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. Continued exposure to depression increases the risk of metabolic disease, cognitive decline, and death\u003csup\u003e\u003cspan additionalcitationids=\"CR7\" citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eWe previously conceptualized recurrence of LLD through a model of neural network homeostatic dysregulation\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Dynamic neural circuitry copes with stressors through allostatic responses encompassing both behavioral and physiologic adaptations\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. While remission of depressive symptoms may reestablish homeostasis, remitted individuals may show long-term state-independent, altered brain network configurations\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. The homeostatic equilibrium of older remitted depressed individuals may be tenuous and further challenged by the degradation of appropriate allostatic responses in older age. This may render older adults particularly vulnerable to recurrence of depression. Our homeostatic disequilibrium hypothesis of LLD recurrence proposes that vulnerability to new depressive episodes stems from chronic fragility in neural network homeostasis persisting in remission. Stressors evoke aberrant neural responses that disrupt homeostasis, alter network function, and subsequently result in maladaptive cognitive and behavioral activity\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e,\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. This leads to a cycle of further disruption in neural function and, eventually, precipitation of another major depressive episode.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThis model is consistent with recent dynamical systems views of psychiatric disorders\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e,\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e, which postulates depression and health as attractor states. The attractor landscape defines the likelihood of relapse (conceptualized as critical transition), though the landscape is defined in abstract terms. Depression has been conceptualized as a disruption of large-scale brain networks, particularly the default mode network (DMN), executive control network (ECN), and salience network (SN)\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e,\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e,\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. These networks are altered in both aging and depression\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. More recently, the somatomotor network (SMN) and visual network have also been implicated in depression and specifically LLD\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e,\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. These networks are typically characterized with resting state functional magnetic resonance imaging (fMRI) studies. In addition to evidence for alterations in depression, we and others have demonstrated that, while resting state functional connectivity (FC) is affected by antidepressant treatment\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e, alterations persist in remission compared to healthy comparison older adults\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e. Thus, FC within and between large scale intrinsic brain networks may comprise a meaningful characterization of the neural landscape that define susceptibility or resilience to critical transitions leading to relapse.\u003c/p\u003e \u003cp\u003eIn this study, we aimed to identify neurobiological factors associated with LLD recurrence risk by conducting a two-year longitudinal study of remitted LLD participants and health comparison older adults. This is one of the first studies to prospectively assess neural markers of relapse and recurrence in depression\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e and, to our knowledge, the first to do so in LLD. For this analysis, we compared resting state FC of seven canonical intrinsic brain networks in older healthy comparisons and participants recently remitted form an acute depressive episode. While our homeostatic disequilibrium model is primarily concerned with dynamic evolution of depressive symptoms and neural architecture, we can use the model to make specific predictions about how stable remission and relapse may differ from each other and healthy comparisons. Specifically, we would expect healthy comparisons to model a stable landscape. Stable remission should present as a similarly stable landscape, while the landscape of eventual relapse should present with greater differences that render it unstable. However, we also expect differences between the remitted landscape, regardless of eventual relapse status, and healthy comparisons. We hypothesized that: 1) within and between network FC, which we define as the neural landscape, will differ between healthy comparisons and remitted participants (regardless of eventual relapse status); 2) remitted participants who relapse will exhibit a different neural landscape compared to participants with stable remission, reflecting unstable vs. stable homeostatic setpoints; 3) the neural landscape in stable remission will be more similar to healthy comparisons than relapse will; and 4) these differences will largely lie in the DMN, ECN, and SN.\u003c/p\u003e"},{"header":"2. Results","content":"\u003cp\u003eDemographic and clinical characteristics are provided in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. Three participants did not have MRI data, and 31 participants had excessive motion (greater than 20% of volumes had root mean square motion\u0026thinsp;\u0026gt;\u0026thinsp;0.5 mm) during the resting state scans, yielding a sample of 111 participants (39 healthy comparison, 47 stable remitted, 25 relapsed) for analysis. The group composition differed by site, with significantly more LLD participants recruited at VUMC due to differing COVID-19 restrictions at Pitt and UIC early in the study. Sex and education differed between groups and was controlled for in all models along with age, race, and site.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eParticipant summary. ITP \u0026ndash; initial treatment phase; LLD \u0026ndash; late-life depression; MADRS \u0026ndash; Montgomery Asberg Depression Rating Scale; SD \u0026ndash; standard deviation.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eOverall\u003c/p\u003e \u003cp\u003e(n\u0026thinsp;=\u0026thinsp;145)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eHealthy comparison\u003c/p\u003e \u003cp\u003e(n\u0026thinsp;=\u0026thinsp;43)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eLLD\u003c/p\u003e \u003cp\u003e(n\u0026thinsp;=\u0026thinsp;102)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eDifference,\u003c/p\u003e \u003cp\u003e\u003cem\u003ep\u003c/em\u003e value\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"1\" nameend=\"c6\" namest=\"c6\"\u003e\u0026nbsp;\u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eAge (mean, SD)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e66.9 (4.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e66.6 (5.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e67.0 (4.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{t}_{68}=-0.35,\\:p=0.723\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"1\" nameend=\"c6\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eSex (female, %)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e95 (65.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e21 (48.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e74 (72.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\chi\\:}_{1}^{2}=5.31,\\:p=0.021\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"1\" nameend=\"c6\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eRace\u003c/b\u003e\u003csup\u003e\u003cb\u003e2\u003c/b\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003e\u003cb\u003eWhite\u003c/b\u003e\u003c/p\u003e \u003cp\u003e\u003cb\u003eBlack\u003c/b\u003e\u003c/p\u003e \u003cp\u003e\u003cb\u003eAsian\u003c/b\u003e\u003c/p\u003e \u003cp\u003e\u003cb\u003eNative Am.\u003c/b\u003e\u003c/p\u003e \u003cp\u003e\u003cb\u003eMultiracial\u003c/b\u003e\u003c/p\u003e \u003cp\u003e\u003cb\u003eOther (7)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e119 (82.1%)\u003c/p\u003e \u003cp\u003e15 (10.3%)\u003c/p\u003e \u003cp\u003e4 (2.8%)\u003c/p\u003e \u003cp\u003e1 (0.7%)\u003c/p\u003e \u003cp\u003e5 (3.4%)\u003c/p\u003e \u003cp\u003e1 (0.7%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e31 (72.1%)\u003c/p\u003e \u003cp\u003e6 (14.0%)\u003c/p\u003e \u003cp\u003e4 (9.3%)\u003c/p\u003e \u003cp\u003e1 (2.3%)\u003c/p\u003e \u003cp\u003e1 (2.3%)\u003c/p\u003e \u003cp\u003e0 (0.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e88 (86.3%)\u003c/p\u003e \u003cp\u003e9 (8.8%)\u003c/p\u003e \u003cp\u003e0 (0.0%)\u003c/p\u003e \u003cp\u003e0 (0.0%)\u003c/p\u003e \u003cp\u003e0 (0.0%)\u003c/p\u003e \u003cp\u003e1 (1.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\chi\\:}_{1}^{2}=3.23,\\:p=0.072\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"1\" nameend=\"c6\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eEducation\u003c/b\u003e\u003c/p\u003e \u003cp\u003e\u003cb\u003e(mean, SD)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e16.1 (2.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e16.7 (2.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e15.8 (2.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{t}_{94}=2.19,\\:p=0.031\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"1\" nameend=\"c6\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eSite\u003c/b\u003e\u003c/p\u003e \u003cp\u003e\u003cb\u003eVUMC\u003c/b\u003e\u003c/p\u003e \u003cp\u003e\u003cb\u003ePitt\u003c/b\u003e\u003c/p\u003e \u003cp\u003e\u003cb\u003eUIC\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e77 (53.1%)\u003c/p\u003e \u003cp\u003e34 (23.4%)\u003c/p\u003e \u003cp\u003e34 (23.4%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e14 (32.6%)\u003c/p\u003e \u003cp\u003e14 (32.6%)\u003c/p\u003e \u003cp\u003e15 (34.9%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e63 (61.8%)\u003c/p\u003e \u003cp\u003e20 (19.6%)\u003c/p\u003e \u003cp\u003e19 (18.6%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\chi\\:}_{2}^{2}=10.43,\\:p=0.005\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"1\" nameend=\"c6\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eBaseline MADRS\u003c/b\u003e\u003c/p\u003e \u003cp\u003e\u003cb\u003e(mean, SD)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4.1 (3.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.9 (1.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5.5 (3.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{t}_{139}=11.35,\\:p\u0026lt;0.001\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"1\" nameend=\"c6\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eTreatment type\u003c/b\u003e\u003c/p\u003e \u003cp\u003e\u003cb\u003e(ITP, %)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e58 (57%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"1\" nameend=\"c6\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eHealthy comparison\u003c/b\u003e\u003c/p\u003e \u003cp\u003e\u003cb\u003e(n\u0026thinsp;=\u0026thinsp;43)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003eStable Remitted\u003c/b\u003e\u003c/p\u003e \u003cp\u003e\u003cb\u003e(n\u0026thinsp;=\u0026thinsp;59)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003eRelapsed\u003c/b\u003e\u003c/p\u003e \u003cp\u003e\u003cb\u003e(n\u0026thinsp;=\u0026thinsp;43)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cb\u003eDifference\u003c/b\u003e,\u003c/p\u003e \u003cp\u003e\u003cb\u003ep\u003c/b\u003e \u003cb\u003evalue\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"1\" nameend=\"c6\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eAge (mean, SD)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e66.6 (5.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e66.9 (4.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e67.0 (4.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{F}_{2}=0.07,\\:p=0.930\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"1\" nameend=\"c6\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eSex\u003c/b\u003e\u003csup\u003e\u003cb\u003ea\u003c/b\u003e\u003c/sup\u003e \u003cb\u003e(female, %)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e21 (48.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e41 (69.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e33 (76.7%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\chi\\:}_{2}^{2}=8.11,\\:p=0.017\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"1\" nameend=\"c6\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eRace\u003c/b\u003e\u003csup\u003e\u003cb\u003eb\u003c/b\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003e\u003cb\u003eWhite\u003c/b\u003e\u003c/p\u003e \u003cp\u003e\u003cb\u003eBlack\u003c/b\u003e\u003c/p\u003e \u003cp\u003e\u003cb\u003eAsian\u003c/b\u003e\u003c/p\u003e \u003cp\u003e\u003cb\u003eNative Am.\u003c/b\u003e\u003c/p\u003e \u003cp\u003e\u003cb\u003eMultiracial\u003c/b\u003e\u003c/p\u003e \u003cp\u003e\u003cb\u003eOther (7)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e31 (72.1%)\u003c/p\u003e \u003cp\u003e6 (14.0%)\u003c/p\u003e \u003cp\u003e4 (9.3%)\u003c/p\u003e \u003cp\u003e1 (2.3%)\u003c/p\u003e \u003cp\u003e1 (2.3%)\u003c/p\u003e \u003cp\u003e0 (0.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e51 (86.4%)\u003c/p\u003e \u003cp\u003e5 (4.9%)\u003c/p\u003e \u003cp\u003e0 (0.0%)\u003c/p\u003e \u003cp\u003e0 (0.0%)\u003c/p\u003e \u003cp\u003e2 (3.4%)\u003c/p\u003e \u003cp\u003e1 (1.7%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e37 (86.0%)\u003c/p\u003e \u003cp\u003e4 (9.3%)\u003c/p\u003e \u003cp\u003e0 (0.0%)\u003c/p\u003e \u003cp\u003e0 (0.0%)\u003c/p\u003e \u003cp\u003e2 (4.7%)\u003c/p\u003e \u003cp\u003e0 (0.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\chi\\:}_{2}^{2}=4.14,\\:p=0.126\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"1\" nameend=\"c6\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eEducation\u003c/b\u003e\u003c/p\u003e \u003cp\u003e\u003cb\u003e(mean, SD)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e16.7 (2.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e15.9 (2.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e15.6 (2.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{F}_{2}=2.32,\\:p=0.102\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"1\" nameend=\"c6\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eSite\u003c/b\u003e\u003c/p\u003e \u003cp\u003e\u003cb\u003eVUMC\u003c/b\u003e\u003c/p\u003e \u003cp\u003e\u003cb\u003ePitt\u003c/b\u003e\u003c/p\u003e \u003cp\u003e\u003cb\u003eUIC\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e14 (32.6%)\u003c/p\u003e \u003cp\u003e14 (32.6%)\u003c/p\u003e \u003cp\u003e15 (34.9%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e31 (52.5%)\u003c/p\u003e \u003cp\u003e14 (23.7%)\u003c/p\u003e \u003cp\u003e14 (23.7%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e32 (74.4%)\u003c/p\u003e \u003cp\u003e6 (14.0%)\u003c/p\u003e \u003cp\u003e5 (11.6%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colspan=\"1\" nameend=\"c6\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eBaseline MADRS\u003c/b\u003e\u003c/p\u003e \u003cp\u003e\u003cb\u003e(mean, SD)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.9 (1.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5.4 (3.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e5.7 (3.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{F}_{2}=38.01,\\:p\u0026lt;0.001\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eTreatment type\u003c/b\u003e\u003c/p\u003e \u003cp\u003e\u003cb\u003e(ITP, %)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e30 (50.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e28 (65.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\chi\\:}_{1}^{2}=4.17,\\:p=0.041\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"1\" nameend=\"c6\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"6\"\u003ea. Self-reported\u003c/td\u003e\u003c/tr\u003e \u003ctr\u003e\u003ctd colspan=\"6\"\u003eb. Race coded as binary variable for analysis with groups of black/Asian/Native American/Multiracial/Other and white.\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1. All remitted LLD participants vs. healthy comparisons\u003c/h2\u003e \u003cp\u003e \u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eResults are summarized in\u003c/span\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e \u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eand full detail is provided in supplemental Table\u0026nbsp;1. LLD participants exhibited lower FC than healthy comparisons within the DMN (\u003c/span\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=7.94,\\:p=0.0008\\)\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e), SN (\u003c/span\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=10.4,\\:p=0.0001\\)\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e), visual (\u003c/span\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=7.18,p=0.002\\)\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e), and SMN (\u003c/span\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=20.0,\\:p\u0026lt;0.0001\\)\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e), as well as between the visual network and DMN (\u003c/span\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=9.16,p=0.0002\\)\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e), limbic (\u003c/span\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=7.86,p=0.0008\\)\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e), and SMN (\u003c/span\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=18.9,p\u0026lt;0.0001\\)\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e). FC between the DMN and ECN was higher in LLD participants than healthy comparisons (\u003c/span\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=12.8,p\u0026lt;0.0001\\)\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e).\u003c/span\u003e\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2. Stable remitted vs. relapse vs. healthy comparisons\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eResults are summarized in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e and full detail is provided in supplemental Table\u0026nbsp;2. For the three group comparison, the DMN exhibited significant FC differences between all 7 networks: ECN (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=10.2,p=0.0001\\)\u003c/span\u003e\u003c/span\u003e), SN (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=9.43,p=0.0001\\)\u003c/span\u003e\u003c/span\u003e), dorsal attention network (DAN) (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=12.6,p\u0026lt;0.0001\\)\u003c/span\u003e\u003c/span\u003e), limbic (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=7.93,p=0.0008\\)\u003c/span\u003e\u003c/span\u003e), visual (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=11.5,p\u0026lt;0.0001\\)\u003c/span\u003e\u003c/span\u003e), SMN (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=9.24,p=0.0002\\)\u003c/span\u003e\u003c/span\u003e) and within DMN (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=10.4,p=0.0001\\)\u003c/span\u003e\u003c/span\u003e). Additional differences were observed between the ECN and visual (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=10.7,p\u0026lt;0.0001\\)\u003c/span\u003e\u003c/span\u003e) and SMN (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=8.00,p=0.0007\\)\u003c/span\u003e\u003c/span\u003e), between the DAN and SMN (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=7.23,p=0.0017\\)\u003c/span\u003e\u003c/span\u003e), and between visual and SMN (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=7.43,p=0.0015\\)\u003c/span\u003e\u003c/span\u003e). Only these network pairs were tested for pairwise group differences to determine which groups exhibited differences and directionality of effects.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cdiv id=\"Sec5\" class=\"Section3\"\u003e \u003ch2\u003e2.2.1. Stable remitted vs. relapse participants\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eResults are summarized in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA and full detail is provided in supplemental Table\u0026nbsp;3. LLD participants who went on to relapse exhibited lower FC than stable remitted participants between the DMN and SN (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=12.9,p\u0026lt;0.0001\\)\u003c/span\u003e\u003c/span\u003e), DMN and SMN (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=17.0,p\u0026lt;0.0001\\)\u003c/span\u003e\u003c/span\u003e), and ECN and SMN (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=7.67,p=0.0014\\)\u003c/span\u003e\u003c/span\u003e). The DMN-SN (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=13.6,p\u0026lt;0.0001)\\)\u003c/span\u003e\u003c/span\u003e and DMN-SMN (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=19.3,p\u0026lt;0.0001\\)\u003c/span\u003e\u003c/span\u003e) connectivities were also indicative of time to relapse in the survival analysis.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section3\"\u003e \u003ch2\u003e2.2.2. Stable remitted and relapse vs. healthy comparisons\u003c/h2\u003e \u003cp\u003eResults are summarized in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB and \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eC and full detail is provided in supplemental Tables\u0026nbsp;4 and 5. Stable remitted participants exhibited widespread differences from healthy comparisons, including greater FC between DMN and ECN (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=15.7,p\u0026lt;0.0001\\)\u003c/span\u003e\u003c/span\u003e), SN(\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=9.18,p=0.0002\\)\u003c/span\u003e\u003c/span\u003e), and SMN (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=10.3,p=0.0001\\)\u003c/span\u003e\u003c/span\u003e) and lower FC within the DMN (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=7.21,p=0.0017\\)\u003c/span\u003e\u003c/span\u003e), between DMN and visual (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=12.2,p\u0026lt;0.0001\\)\u003c/span\u003e\u003c/span\u003e), and between visual and SMN (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=19.9,p\u0026lt;0.0001\\)\u003c/span\u003e\u003c/span\u003e). By comparison, participants who go on to relapse have far fewer differences from healthy comparisons: greater FC between DMN and ECN (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=9.64,p=0.0001\\)\u003c/span\u003e\u003c/span\u003e) and lower FC between SMN and DMN (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=8.87,p=0.0003\\)\u003c/span\u003e\u003c/span\u003e) and visual (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:HC=17.2,p\u0026lt;0.0001\\)\u003c/span\u003e\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eTo test this further, we computed participant-wise similarity matrices for the functional connectomes and compared the similarities between healthy comparison and stable remitted to the similarities between healthy comparison and relapse. Overall, the neural landscape of the relapse group was more similar to healthy comparisons than the neural landscape of the stable remission group was (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:t=-2.51,p=0.0123\\)\u003c/span\u003e\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"3. Discussion","content":"\u003cp\u003eIn this study of relapse and recurrence in late-life depression, we found robust differences in network connectivity between relapse, stable remission, and heathy comparisons, differences that centered heavily on the DMN. The remitted LLD group as a whole exhibited lower within and between network connectivity of the DMN, visual, and SMN than healthy comparisons, with the exception of greater connectivity between DMN and ECN. Stable remission and relapse differed primarily by lower connectivity between DMN and SN and SMN in the relapse group. Overall, the connectivity of participants that went on to relapse was more similar to healthy comparisons than was the connectivity of the stable remitted participants.\u003c/p\u003e \u003cp\u003eConsistent with our first hypothesis, the remitted LLD group as a whole showed significant differences in network connectivity from healthy comparisons, reflecting a new homeostatic setpoint associated with remission. The DMN featured prominently in these differences, consistent with previous working implicating aberrant DMN connectivity in both adult and geriatric depression\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e,\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. The finding of greater connectivity between the DMN and ECN in the remitted LLD group partially replicates findings in younger adults that connectivity between specific DMN-ECN regions was elevated in relapsing participants (though not in stable remitted participants)\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e,\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e. There is also evidence that DMN-ECN connectivity is higher in acutely depressed individuals\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e, while our previous review identified a positive association between DMN-ECN connectivity and antidepressant treatment response\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e. The coherence of the DMN and ECN appears to be a key feature of depression, though the precise role remains unclear. The SMN and visual network also showed lower within-network connectivity in LLD, which is consistent with findings from a large Chinese consortium\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e,\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e, though at least one small study has reported the opposite effect in the visual network\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. Overall, remitted LLD participants show robust differences from healthy comparisons, supporting that stable remission from depression is not simply a \u0026ldquo;return to normal,\u0026rdquo; but is associated with a new configuration of the neural landscape.\u003c/p\u003e \u003cp\u003eThe neural landscape also differed between stabled remitted and relapsed participants. DMN-SN connectivity was lower in relapsed participants than in stable remitted, and was also associated with time to relapse. This is consistent with a recent study in midlife depression that reported relapse was associated with lower connectivity between the DMN and portions of the SN\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e, but stands in contrast to another a recent finding (albeit in only 9 relapsed participants) showing higher FC between the right anterior insula (a SN hub) and the subcallosal cingulate (a DMN hub) predicted depression recurrence\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. Interestingly, lower DMN-SN connectivity has also been reported in acutely depessed individuals\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e. Lower SMN connectivity to the DMN and ECN was also associated with relapse which is the first finding to our knowledge relating SMN connectivity to relapse and requires further exploration.\u003c/p\u003e \u003cp\u003eContrary to our third hypothesis, participants who would go on to relapse displayed a functional connectivity profile more similar to healthy comparisons than the stable remitted participants did. This finding may indicate that significant reconfigurations of the neural landscape in context of successful antidepressant treatment are required to result in stable remission. There is wide-spread evidence for this reconfiguration across the DMN, ECN, and SN\u003csup\u003e\u003cspan additionalcitationids=\"CR29 CR30 CR31\" citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. This type of dynamic network reconfiguration (as opposed to a return to baseline) is consistent with work in learning reversal, showing that new circuits that will override the learned behavior, rather than a reversal of the learned circuitry\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e. As a corollary, a partial return to baseline (i.e., failure to establish a new stable setpoint) may represent an unstable equilibrium and thus susceptibility to relapse, especially in the presence of other persistent alterations. This may be evidenced in our study by relapsed participants showing greater similarity of functional network organization to healthy comparisons than stable remitted participants.\u003c/p\u003e \u003cp\u003eOur final hypothesis\u0026mdash;that the key networks of the LLD neural landscape are the DMN, ECN, and SN\u0026mdash;appears to be mostly true. The DMN showed widespread and persistent differences between the groups, underscoring its crucial role in the neural underpinnings of late-life depression. Importantly, these differences did not lie just within the DMN, but often in the interaction between the DMN and other networks. The ECN and SN did not differ as robustly as the DMN, though their connectivity with the DMN served as important differentiators between healthy comparisons and LLD, and between stable remission and relapse, respectively. Connectivity of the SMN and visual networks also differed frequently between groups, adding to a growing body of evidence that these unimodal networks are also involved in depression\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe prospective design and 4 month inclusion cutoff following remission are significant strengths of our study, providing ecological validity as evidenced by the excellent agreement with previously observed relapse rates of 43% within two years\u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e. This study has several limitations. Our ability to assess the temporal stability of neural networks is hampered by analyzing only baseline neuroimaging. We do not have pretreatment imaging data in the LLD group. Sample size is moderate, although above average for neuroimaging of clinical populations and sufficiently powered for our analytic method that leverages omnibus testing to reduce dimensionality. Remitted participants did not receive uniform antidepressant treatment; while a general treatment algorithm was followed, medications were individually tailored. However, this resulted in a high rate of remission (73% across all three sites) and provides better generalization/clinical translation. The delineation between the stable remitted and relapsed groups was subject to right-censoring; while all participants had\u0026thinsp;\u0026ge;\u0026thinsp;8 months of follow up at the time of analysis, most participants had a longer duration (up to 2 years). We were insufficiently powered to analyze a uniform cutoff of relapse within 8 months, which would result in only 13 relapsed participants. Since relapse tends to occur sooner rather than later (e.g., only 29% of relapsed participants with two years of data relapsed after one year) and 85% of the participants had\u0026thinsp;\u0026ge;\u0026thinsp;1 year of follow up, we believe our inclusive approach represents a better approximation to the \u0026ldquo;true\u0026rdquo; relapse group that we would observe with 2 years of longitudinal data for all participants. In this study we chose to define the neural landscape in terms of within and between network connectivity using the canonical seven Yeo networks. While this choice is well-justified, there are models/evidence that support other meaningful organizational levels of inquiry (e.g., subnetworks or specific regions, especially subcortical regions). Restricting our analysis to 7 networks allowed for cortex-wide coverage while also minimizing the number of comparisons. This investigation is focused on resting state fMRI; future investigations will tests the effect of task-based fMRI or structural connectivity (e.g., white matter hyperintensities, diffusion imagining measures).\u003c/p\u003e \u003cp\u003eIn conclusion, we identified robust differences in the functional connectome between healthy comparison and remitted participants in late-life, as well as differences between participants who relapsed and those who remained depression-free. These differences were apparent at the network level and were robust to parcellation scheme. These findings, consistent with our proposed disruption of neural homeostasis model of recurrence and relapse in late-depression, may be used to identify depressed older adults at higher risk of relapse and to adequately tailor preventative interventions to prevent the burden of additional depressive episodes.\u003c/p\u003e"},{"header":"4. Methods","content":"\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e4.1. Overview\u003c/h2\u003e \u003cp\u003e Participants were enrolled in the multisite REMBRANDT Study (Recurrence markers, cognitive burden, and neurobiological homeostasis in late-life depression) at Vanderbilt University Medical Center (VUMC), University of Pittsburgh (Pitt.), and University of Illinois \u0026ndash; Chicago (UIC). A full description of the study design and rationale has been published previously\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e. Briefly, LLD participants enter through an Initial Treatment Phase (ITP) if currently depressed, or are recruited directly into the longitudinal phase if they recently (\u0026lt;\u0026thinsp;4 months) remitted to clinical treatment. LLD participants were treated for up to 20 weeks with an algorithm informed by STAR*D\u003csup\u003e36\u003c/sup\u003e and the Duke Neurocognitive Outcomes of Depression in the Elderly studies\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e. Individuals who did not remit were referred for clinical care. Remission was defined as Montgomery-\u0026Aring;sberg Depression Rating Scale\u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e (MADRS) score of \u0026le;\u0026thinsp;10 concluding the ITP and longitudinal baseline visit, which must occur within 4 months (but no sooner than 1 month) following remission. Remitted LLD and healthy comparison participants entered the two-year longitudinal phase involving scheduled contact every 2 months. This study was approved by the institutional review boards at all three sites. All participants provided informed consent. A CONSORT diagram for the study is shown in Supplemental Fig.\u0026nbsp;1.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e4.2. Participants\u003c/h2\u003e \u003cp\u003eWe enrolled 145 participants, including 102 remitted LLD participants and 43 healthy comparison participants who completed baseline neuroimaging and at least 8 months of clinical follow-up. Inclusion criteria for all participants were: age\u0026thinsp;\u0026ge;\u0026thinsp;60, fluent in English, Montreal Cognitive Assessment (MoCA)\u0026thinsp;\u0026ge;\u0026thinsp;24 or MoCA-BLIND\u0026thinsp;\u0026ge;\u0026thinsp;18. Further inclusion criteria for LLD: current diagnosis of recurrent major depressive disorder (MDD) based on the Diagnostic and Statistical Manual of Mental Disorders (DSM-5), and MADRS\u0026thinsp;\u0026ge;\u0026thinsp;15 for depressed participants entering the ITP; remission from current MDD episode and MADRS\u0026thinsp;\u0026le;\u0026thinsp;10 for LLD participants entering the longitudinal phase. Exclusion criteria for all participants: current or past Axis I diagnosis except generalized anxiety disorder, panic disorder, or simple phobias; history of substance dependence/abuse in the past year; acute grief or suicidality; current or past psychosis; neurological disorders, including dementia; unstable medical illness requiring treatment; MRI contraindications, electroconvulsive therapy in the past 6 months; current brain stimulation or ketamine/esketamine treatment. Healthy comparison participants were excluded for MADRS scores\u0026thinsp;\u0026gt;\u0026thinsp;8, current or past depression diagnosis or use of psychotropic medication for psychiatric symptoms. Past brief therapy for specific challenges or losses was allowable.\u003c/p\u003e \u003cp\u003e We collected demographic, clinical, neuropsychological and MRI data from all participants. While data collection is currently ongoing, we have at least 8 months of longitudinal clinical data on all participants. Based on the available longitudinal data, we further stratified the LLD group into 59 stable remitted participants and 43 relapsed participants (see Assessments for definition).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e4.3. Assessments\u003c/h2\u003e \u003cp\u003eDepression severity was assessed at baseline and every two months throughout the study. Remitted LLD participants who maintained MADRS\u0026thinsp;\u0026le;\u0026thinsp;15 over the duration of the study were classified as \u003cspan type=\"BoldItalicUnderline\" class=\"BoldItalicUnderline\" name=\"Emphasis\"\u003estable remitted\u003c/span\u003e. Initially remitted participants who experienced a MADRS\u0026thinsp;\u0026gt;\u0026thinsp;15 plus DSM-5 MDD criteria for at least two weeks were classified as \u003cspan type=\"BoldItalicUnderline\" class=\"BoldItalicUnderline\" name=\"Emphasis\"\u003erelapsed.\u003c/span\u003e In this study, we do not differentiate between relapse (reoccurrence of symptoms from previous episode) and recurrence (occurrence of a new depressive episode) since this delineation is often unclear or relies on arbitray time cutoffs.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e4.4. MRI Acquisition\u003c/h2\u003e \u003cp\u003eMRI were acquired 1\u0026ndash;4 months after remission from an acute depressive episode at 3T on Philips Elition (VUMC), Siemens Prisma (Pitt) and GE Discovery MR750 System (UIC) scanners using 32-channel head coils following the Adolescent Brain Cognitive Development (ABCD) Study MRI Protocol\u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e to harmonize across sites. We collected high resolution structural images with a sagittal T1-weighted Magnetization Prepared Rapid Gradient Echo (MPRAGE) sequence, an axial T2-weighted Fluid Attenuated Inversion Recovery (FLAIR) sequence, resting state with an axial T2* blood oxygen level dependent (BOLD) gradient-echo echoplanar imaging sequence, and a reverse encoded top-up image for correcting susceptibility artifacts matched to the resting sate acquisition parameters. Specific acquisition parameters are provided in Supplemental Tables\u0026nbsp;4 and 5 of the study protocol\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e. Two separate 5 minute resting state scans (188 volumes per scan) were acquired back-to-back, which has been shown to improve reliability of measures\u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e. We assessed scanner variability across the sites using structural\u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e and functional phantoms\u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e. All sites showed acceptable levels of signal-to-noise, though there were differences by site; therefore, we control for site in all analyses.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e4.5. MRI Processing\u003c/h2\u003e \u003cp\u003eMRI were processed at VUMC using Docker containers in XNAT to ensure reproducibility with in-house scripts utilizing the Statistical Parametric Mapping (SPM12) toolbox\u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e except where noted. The FLAIR image was coregistered to the MPRAGE and both images were used in multispectral segmentation into six canonical tissue types, which produces a deformation field for normalizing images into Montreal Neurological Institute (MNI) space. Two were Gaussians used for white matter to account for white matter hyperintensities common in older adults\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e. The resulting grey matter, white matter, and cerebrospinal fluid (CSF) tissue maps were thresholded at 0.1, filled, and smoothed to create an intracranial volume mask used to skull-strip the structural images.\u003c/p\u003e \u003cp\u003eResting state images were slice-time corrected, corrected for susceptibility induced distortions using topup in FSL\u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e, realigned to the mean volume using rigid body transformations, skull-stripped using the brain extraction tool in FSL\u003csup\u003e\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e, coregistered to the MPRAGE, normalized to MNI space (2mm\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e isotropic resolution) using the structural deformation field, and smoothed with a Gaussian kernel with 8mm full width at half maximum. The first five principal components of the white matter and CSF, six motion parameters, and sinusoids for bandpass filtering in the 0.008\u0026ndash;0.15 Hz frequency range were regressed out of the image.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003e4.6. Connectome Calculation\u003c/h2\u003e \u003cp\u003eWe calculated FC matrices by extracting the first principal component of processed image time series in predefined parcellation regions and quantifying region-to-region FC as Pearson correlation of the time series. To ensure results are robust to parcellation scheme, we employed three different functionally-defined parcellations: Shen268\u003csup\u003e47\u003c/sup\u003e, Schaefer200, and Schaefer400\u003csup\u003e48\u003c/sup\u003e, resulting in three FC matrices of size 268x268, 200x200, and 400x400 per scan, with each entry referred to as an edge in the connectome. We overlayed the Yeo 7 Network definitions\u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e (DMN, ECN, SN, DAN, limbic network, visual network, and SMN) on these atlases and assigned each node to a canonical intrinsic network with a winner-take-all approach. We restrict our network analysis to within and between network connectivity of these 7 networks for a total of 28 \u0026ldquo;features\u0026rdquo; of the neural landscape.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e4.7. Statistical Analysis\u003c/h2\u003e \u003cp\u003eWe assessed for differences between the healthy comparison and LLD participants (2 group comparison) and between the healthy comparison, stable remitted, and relapsed groups (3 group comparison) using the generalized linear model. Post-hoc tests for significant 3 group comparison results are performed to determine which groups show differences and the directionality of the effect. We also tested time-to-relapse in the LLD group using a Cox proportional hazards model. Age, sex, race, education, and site were controlled for in all models.\u003c/p\u003e \u003cp\u003eFor this analysis, we employ mass univariate testing at the edge-level (each entry of the FC matrices) and perform inference at higher organizational levels (e.g., network) on subsets of these tests using higher criticism (HC). HC is omnibus test optimal for detecting rare and weak signals\u003csup\u003e\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e,\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e. This is accomplished by comparing the observed \u003cem\u003ep\u003c/em\u003e-values to the theoretical null using a modified Kolmogrov-Smirnov statistic; hence \u003cem\u003eHC only tests for difference among a collection of edges\u003c/em\u003e\u0026mdash;\u003cem\u003eit does NOT offer inference at the edge-level.\u003c/em\u003e This approach optimally balances power and specificity for connectome analysis and has recently been applied to neuroimaging\u003csup\u003e\u003cspan additionalcitationids=\"CR53\" citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e\u003c/sup\u003e.At the network level, HC is applied only to entries in the FC matrix corresponding to the specified network pairs (i.e., partitioned submatrices for between network connectivity and triangular blocks on the diagonal for within network connectivity). Thus, while we preform primary statistical tests on full connectome matrices with 200 to 400 nodes, we only perform inference on the 7-network connectome with the many region-to-region tests essentially functioning as multiple observations within each network pair. We apply the Benjamini-Hochberg procedure for controlling the False Discovery Rate at \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\alpha\\:=0.05\\)\u003c/span\u003e\u003c/span\u003e with \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:m=28\\)\u003c/span\u003e\u003c/span\u003e. We calculate \u003cem\u003ep\u003c/em\u003e values for the HC statistic by comparison to a Monte Carlo simulation of 10,000 null distributions for the given number of tests (edges or entries in the FC submatrix).\u003c/p\u003e \u003cp\u003eTo compare the connectivity profiles of stable remitted and relapse participants to healthy comparisons, we computed a participant similarity matrix. Similarity was calculated as the inverse of the Euclidean distance between the FC matrix entries of two participants (inverse of the Frobenius norm of the difference FC matrix). Similarity between healthy comparisons and stable remitted was compared to similarity between healthy comparisons and relapse with \u003cem\u003et\u003c/em\u003e-tests. Note that age, sex, race, and education were not controlled for in this analysis since each similarity metric compares two participants.\u003c/p\u003e \u003cp\u003eWe only report network results that were significant across at least two of the parcellations and note results that were significant across all three parcellations. For brevity, we only report median test statistics across the three parcellations in the text with results for all parcellations contained in the supplement.\u003c/p\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eDisclosures:\u0026nbsp;\u003c/strong\u003eOlusola Ajilore is a co-founder of Keywise AI, has served as a consultant for Sage Therapeutics and Otsuka, has received honoraria from Boehringer Ingelheim, and is on the advisory board for Blueprint Health and Embodied Labs. Antonija Kolobaric serves as a consultant for Radicle Science. No other authors have disclosures to report.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements:\u0026nbsp;\u003c/strong\u003eWe\u0026rsquo;d like to thank the numerous staff at VUMC, Pitt, and UIC who helped administer the study protocol and process the data.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePrevious presentation:\u003c/strong\u003e This work was presented as a poster at conferences for the American Association for Geriatric Psychiatry (March 16, 2024, Atlanta, GA) and the Society for Biological Psychiatry (May 10, 2024, Austin, TX).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Availability:\u0026nbsp;\u003c/strong\u003eDeidentified data is available upon reasonable request. The analysis script is available at https://github.com/REMBRANDT-study/resting_state_network_analysis_baseline145.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding:\u003c/strong\u003e This work was supported by National of Institute of Health grants R01 MH121619, R01 MH121620, R01 MH121384, K01 MH133913, R01 MH108509, K01 MH122741, T32 MH019986, and the National Center for Advancing Translational Sciences grants\u0026nbsp;UL1 TR000445\u0026nbsp;and UL1 TR002243.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cbr\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eTaylor Warren D. Depression in the Elderly. \u003cem\u003eN Engl J Med\u003c/em\u003e. 2014;371(13):1228\u0026ndash;1236. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1056/NEJMcp1402180\u003c/span\u003e\u003cspan address=\"10.1056/NEJMcp1402180\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSzymkowicz SM, Gerlach AR, Homiack D, Taylor WD. Biological factors influencing depression in later life: role of aging processes and treatment implications. \u003cem\u003eTransl Psychiatry\u003c/em\u003e. 2023;13(1):1\u0026ndash;16. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41398-023-02464-9\u003c/span\u003e\u003cspan address=\"10.1038/s41398-023-02464-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePenninx BWJH, Geerlings SW, Deeg DJH, van Eijk JTM, van Tilburg W, Beekman ATF. Minor and Major Depression and the Risk of Death in Older Persons. \u003cem\u003eArch Gen Psychiatry\u003c/em\u003e. 1999;56(10):889\u0026ndash;895. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1001/archpsyc.56.10.889\u003c/span\u003e\u003cspan address=\"10.1001/archpsyc.56.10.889\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBeekman ATF, Geerlings SW, Deeg DJH, et al. The Natural History of Late-Life Depression: A 6-Year Prospective Study in the Community. \u003cem\u003eArch Gen Psychiatry\u003c/em\u003e. 2002;59(7):605\u0026ndash;611. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1001/archpsyc.59.7.605\u003c/span\u003e\u003cspan address=\"10.1001/archpsyc.59.7.605\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eReynolds CF, Dew MA, Pollock BG, et al. Maintenance treatment of major depression in old age. \u003cem\u003eN Engl J Med\u003c/em\u003e. 2006;354(11):1130\u0026ndash;1138. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1056/NEJMoa052619\u003c/span\u003e\u003cspan address=\"10.1056/NEJMoa052619\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePenninx BWJH. Depression and cardiovascular disease: Epidemiological evidence on their linking mechanisms. \u003cem\u003eNeurosci Biobehav Rev\u003c/em\u003e. 2017;74:277\u0026ndash;286. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.neubiorev.2016.07.003\u003c/span\u003e\u003cspan address=\"10.1016/j.neubiorev.2016.07.003\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePan A, Sun Q, Okereke OI, Rexrode KM, Hu FB. Depression and Risk of Stroke Morbidity and Mortality: A Meta-analysis and Systematic Review. \u003cem\u003eJAMA\u003c/em\u003e. 2011;306(11):1241\u0026ndash;1249. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1001/jama.2011.1282\u003c/span\u003e\u003cspan address=\"10.1001/jama.2011.1282\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGanguli M, Du Y, Dodge HH, Ratcliff GG, Chang CCH. Depressive Symptoms and Cognitive Decline in Late Life: A Prospective Epidemiological Study. \u003cem\u003eArch Gen Psychiatry\u003c/em\u003e. 2006;63(2):153\u0026ndash;160. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1001/archpsyc.63.2.153\u003c/span\u003e\u003cspan address=\"10.1001/archpsyc.63.2.153\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAndreescu C, Ajilore O, Aizenstein HJ, et al. Disruption of Neural Homeostasis as a Model of Relapse and Recurrence in Late-Life Depression. \u003cem\u003eAm J Geriatr Psychiatry\u003c/em\u003e. 2019;27(12):1316\u0026ndash;1330. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.jagp.2019.07.016\u003c/span\u003e\u003cspan address=\"10.1016/j.jagp.2019.07.016\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcEwen BS, Gianaros PJ. Stress- and Allostasis-Induced Brain Plasticity. \u003cem\u003eAnnu Rev Med\u003c/em\u003e. 2011;62(1):431\u0026ndash;445. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1146/annurev-med-052209-100430\u003c/span\u003e\u003cspan address=\"10.1146/annurev-med-052209-100430\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMing Q, Zhong X, Zhang X, et al. State-Independent and Dependent Neural Responses to Psychosocial Stress in Current and Remitted Depression. \u003cem\u003eAm J Psychiatry\u003c/em\u003e. 2017;174(10):971\u0026ndash;979. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1176/appi.ajp.2017.16080974\u003c/span\u003e\u003cspan address=\"10.1176/appi.ajp.2017.16080974\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eScheffer M, Bockting CL, Borsboom D, et al. A Dynamical Systems View of Psychiatric Disorders\u0026mdash;Theory: A Review. \u003cem\u003eJAMA Psychiatry\u003c/em\u003e. Published online April 3, 2024. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1001/jamapsychiatry.2024.0215\u003c/span\u003e\u003cspan address=\"10.1001/jamapsychiatry.2024.0215\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eScheffer M, Bockting CL, Borsboom D, et al. A Dynamical Systems View of Psychiatric Disorders\u0026mdash;Practical Implications: A Review. \u003cem\u003eJAMA Psychiatry\u003c/em\u003e. Published online April 3, 2024. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1001/jamapsychiatry.2024.0228\u003c/span\u003e\u003cspan address=\"10.1001/jamapsychiatry.2024.0228\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMenon V. Large-scale brain networks and psychopathology: a unifying triple network model. \u003cem\u003eTrends Cogn Sci\u003c/em\u003e. 2011;15(10):483\u0026ndash;506. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.tics.2011.08.003\u003c/span\u003e\u003cspan address=\"10.1016/j.tics.2011.08.003\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGunning FM, Oberlin LE, Schier M, Victoria LW. Brain-based mechanisms of late-life depression: Implications for novel interventions. \u003cem\u003eSemin Cell Dev Biol\u003c/em\u003e. 2021;116:169\u0026ndash;179. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.semcdb.2021.05.002\u003c/span\u003e\u003cspan address=\"10.1016/j.semcdb.2021.05.002\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTan W, Ouyang X, Huang D, et al. Disrupted intrinsic functional brain network in patients with late-life depression: Evidence from a multi-site dataset. \u003cem\u003eJ Affect Disord\u003c/em\u003e. 2023;323:631\u0026ndash;639. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.jad.2022.12.019\u003c/span\u003e\u003cspan address=\"10.1016/j.jad.2022.12.019\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYang H, Chen X, Chen ZB, et al. Disrupted intrinsic functional brain topology in patients with major depressive disorder. \u003cem\u003eMol Psychiatry\u003c/em\u003e. 2021;26(12):7363\u0026ndash;7371. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41380-021-01247-2\u003c/span\u003e\u003cspan address=\"10.1038/s41380-021-01247-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKarim HT, Andreescu C, Tudorascu D, et al. Intrinsic functional connectivity in late-life depression: trajectories over the course of pharmacotherapy in remitters and non-remitters. \u003cem\u003eMol Psychiatry\u003c/em\u003e. 2017;22(3):450\u0026ndash;457. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/mp.2016.55\u003c/span\u003e\u003cspan address=\"10.1038/mp.2016.55\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWu M, Andreescu C, Butters MA, Tamburo R, Reynolds CF, Aizenstein H. Default-mode network connectivity and white matter burden in late-life depression. \u003cem\u003ePsychiatry Res Neuroimaging\u003c/em\u003e. 2011;194(1):39\u0026ndash;46. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.pscychresns.2011.04.003\u003c/span\u003e\u003cspan address=\"10.1016/j.pscychresns.2011.04.003\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDunlop BW, Cha J, Choi KS, Nemeroff CB, Craighead WE, Mayberg HS. Functional connectivity of salience and affective networks among remitted depressed patients predicts episode recurrence. \u003cem\u003eNeuropsychopharmacology\u003c/em\u003e. 2023;48(13):1901\u0026ndash;1909. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41386-023-01653-w\u003c/span\u003e\u003cspan address=\"10.1038/s41386-023-01653-w\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKaiser RH, Andrews-Hanna JR, Wager TD, Pizzagalli DA. Large-Scale Network Dysfunction in Major Depressive Disorder: A Meta-analysis of Resting-State Functional Connectivity. \u003cem\u003eJAMA Psychiatry\u003c/em\u003e. 2015;72(6):603\u0026ndash;611. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1001/jamapsychiatry.2015.0071\u003c/span\u003e\u003cspan address=\"10.1001/jamapsychiatry.2015.0071\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLangenecker SA, Jenkins LM, Stange JP, et al. Cognitive control neuroimaging measures differentiate between those with and without future recurrence of depression. \u003cem\u003eNeuroImage Clin\u003c/em\u003e. 2018;20:1001\u0026ndash;1009. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.nicl.2018.10.004\u003c/span\u003e\u003cspan address=\"10.1016/j.nicl.2018.10.004\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBerwian IM, Wenzel JG, Kuehn L, et al. The relationship between resting-state functional connectivity, antidepressant discontinuation and depression relapse. \u003cem\u003eSci Rep\u003c/em\u003e. 2020;10:22346. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41598-020-79170-9\u003c/span\u003e\u003cspan address=\"10.1038/s41598-020-79170-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGerlach AR, Karim HT, Peci\u0026ntilde;a M, et al. MRI predictors of pharmacotherapy response in major depressive disorder. \u003cem\u003eNeuroImage Clin\u003c/em\u003e. 2022;36:103157. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.nicl.2022.103157\u003c/span\u003e\u003cspan address=\"10.1016/j.nicl.2022.103157\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEyre HA, Yang H, Leaver AM, et al. Altered resting-state functional connectivity in late-life depression: a cross-sectional study. \u003cem\u003eJ Affect Disord\u003c/em\u003e. 2016;189:126\u0026ndash;133. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.jad.2015.09.011\u003c/span\u003e\u003cspan address=\"10.1016/j.jad.2015.09.011\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan Kleef RS, Kaushik P, Besten M, et al. Understanding and predicting future relapse in depression from resting state functional connectivity and self-referential processing. \u003cem\u003eJ Psychiatr Res\u003c/em\u003e. 2023;165:305\u0026ndash;314. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.jpsychires.2023.07.034\u003c/span\u003e\u003cspan address=\"10.1016/j.jpsychires.2023.07.034\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi G, Liu Y, Zheng Y, et al. Large-scale dynamic causal modeling of major depressive disorder based on resting‐state functional magnetic resonance imaging. \u003cem\u003eHum Brain Mapp\u003c/em\u003e. 2019;41(4):865\u0026ndash;881. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/hbm.24845\u003c/span\u003e\u003cspan address=\"10.1002/hbm.24845\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVega JN, Taylor WD, Gandelman JA, et al. Persistent Intrinsic Functional Network Connectivity Alterations in Middle-Aged and Older Women With Remitted Depression. \u003cem\u003eFront Psychiatry\u003c/em\u003e. 2020;11:62. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fpsyt.2020.00062\u003c/span\u003e\u003cspan address=\"10.3389/fpsyt.2020.00062\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCui J, Wang Y, Liu R, et al. Effects of escitalopram therapy on resting-state functional connectivity of subsystems of the default mode network in unmedicated patients with major depressive disorder. \u003cem\u003eTransl Psychiatry\u003c/em\u003e. 2021;11(1):634. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41398-021-01754-4\u003c/span\u003e\u003cspan address=\"10.1038/s41398-021-01754-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJiao K, Xu H, Teng C, et al. Connectivity patterns of cognitive control network in first episode medication-naive depression and remitted depression. \u003cem\u003eBehav Brain Res\u003c/em\u003e. 2020;379:112381. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.bbr.2019.112381\u003c/span\u003e\u003cspan address=\"10.1016/j.bbr.2019.112381\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWorkman CI, Lythe KE, McKie S, et al. A novel resting-state functional magnetic resonance imaging signature of resilience to recurrent depression. \u003cem\u003ePsychol Med\u003c/em\u003e. 2017;47(4):597\u0026ndash;607. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1017/S0033291716002567\u003c/span\u003e\u003cspan address=\"10.1017/S0033291716002567\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu J, Fan Y, Ling-Li Zeng null, et al. The neuroprogressive nature of major depressive disorder: evidence from an intrinsic connectome analysis. \u003cem\u003eTransl Psychiatry\u003c/em\u003e. 2021;11(1):102. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41398-021-01227-8\u003c/span\u003e\u003cspan address=\"10.1038/s41398-021-01227-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBarron HC. Neural inhibition for continual learning and memory. \u003cem\u003eCurr Opin Neurobiol\u003c/em\u003e. 2021;67:85\u0026ndash;94. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.conb.2020.09.007\u003c/span\u003e\u003cspan address=\"10.1016/j.conb.2020.09.007\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDeng Y, McQuoid DR, Potter GG, et al. Predictors of recurrence in remitted late-life depression. \u003cem\u003eDepress Anxiety\u003c/em\u003e. 2018;35(7):658\u0026ndash;667. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/da.22772\u003c/span\u003e\u003cspan address=\"10.1002/da.22772\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTaylor WD, Ajilore O, Karim HT, et al. Assessing depression recurrence, cognitive burden, and neurobiological homeostasis in late life: Design and rationale of the REMBRANDT study. \u003cem\u003eJ Mood Anxiety Disord\u003c/em\u003e. Published online November 4, 2023:100038. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.xjmad.2023.100038\u003c/span\u003e\u003cspan address=\"10.1016/j.xjmad.2023.100038\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWarden D, Rush AJ, Trivedi MH, Fava M, Wisniewski SR. The STAR*D Project results: a comprehensive review of findings. \u003cem\u003eCurr Psychiatry Rep\u003c/em\u003e. 2007;9(6):449\u0026ndash;459. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s11920-007-0061-3\u003c/span\u003e\u003cspan address=\"10.1007/s11920-007-0061-3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSteffens DC, McQuoid DR, Krishnan KRR. The Duke Somatic Treatment Algorithm for Geriatric Depression (STAGED) approach. \u003cem\u003ePsychopharmacol Bull\u003c/em\u003e. 2002;36(2):58\u0026ndash;68.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMontgomery SA, Asberg M. A new depression scale designed to be sensitive to change. \u003cem\u003eBr J Psychiatry J Ment Sci\u003c/em\u003e. 1979;134:382\u0026ndash;389. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1192/bjp.134.4.382\u003c/span\u003e\u003cspan address=\"10.1192/bjp.134.4.382\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCasey BJ, Cannonier T, Conley MI, et al. The Adolescent Brain Cognitive Development (ABCD) study: Imaging acquisition across 21 sites. \u003cem\u003eAdolesc Brain Cogn Dev ABCD Consort Ration Aims Assess Strategy\u003c/em\u003e. 2018;32:43\u0026ndash;54. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.dcn.2018.03.001\u003c/span\u003e\u003cspan address=\"10.1016/j.dcn.2018.03.001\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCho JW, Korchmaros A, Vogelstein JT, Milham MP, Xu T. Impact of concatenating fMRI data on reliability for functional connectomics. \u003cem\u003eNeuroImage\u003c/em\u003e. 2021;226:117549. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.neuroimage.2020.117549\u003c/span\u003e\u003cspan address=\"10.1016/j.neuroimage.2020.117549\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGunter JL, Bernstein MA, Borowski BJ, et al. Measurement of MRI scanner performance with the ADNI phantom. \u003cem\u003eMed Phys\u003c/em\u003e. 2009;36(6Part1):2193\u0026ndash;2205. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1118/1.3116776\u003c/span\u003e\u003cspan address=\"10.1118/1.3116776\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFriedman L, Glover GH. Report on a multicenter fMRI quality assurance protocol. \u003cem\u003eJ Magn Reson Imaging\u003c/em\u003e. 2006;23(6):827\u0026ndash;839. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/jmri.20583\u003c/span\u003e\u003cspan address=\"10.1002/jmri.20583\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePenny WD, Friston KJ, Ashburner JT, Kiebel SJ, Nichols TE. \u003cem\u003eStatistical Parametric Mapping: The Analysis of Functional Brain Images\u003c/em\u003e. Elsevier; 2011.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKarim HT, Andreescu C, MacCloud RL, et al. The effects of white matter disease on the accuracy of automated segmentation. \u003cem\u003ePsychiatry Res Neuroimaging\u003c/em\u003e. 2016;253:7\u0026ndash;14. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.pscychresns.2016.05.003\u003c/span\u003e\u003cspan address=\"10.1016/j.pscychresns.2016.05.003\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAndersson JLR, Skare S, Ashburner J. How to correct susceptibility distortions in spin-echo echo-planar images: application to diffusion tensor imaging. \u003cem\u003eNeuroImage\u003c/em\u003e. 2003;20(2):870\u0026ndash;888. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/S1053-8119(03)00336-7\u003c/span\u003e\u003cspan address=\"10.1016/S1053-8119(03)00336-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSmith SM. Fast robust automated brain extraction. \u003cem\u003eHum Brain Mapp\u003c/em\u003e. 2002;17(3):143\u0026ndash;155. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/hbm.10062\u003c/span\u003e\u003cspan address=\"10.1002/hbm.10062\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShen X, Tokoglu F, Papademetris X, Constable RT. Groupwise whole-brain parcellation from resting-state fMRI data for network node identification. \u003cem\u003eNeuroImage\u003c/em\u003e. 2013;82:403\u0026ndash;415. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.neuroimage.2013.05.081\u003c/span\u003e\u003cspan address=\"10.1016/j.neuroimage.2013.05.081\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchaefer A, Kong R, Gordon EM, et al. Local-Global Parcellation of the Human Cerebral Cortex from Intrinsic Functional Connectivity MRI. \u003cem\u003eCereb Cortex N Y N 1991\u003c/em\u003e. 2018;28(9):3095\u0026ndash;3114. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/cercor/bhx179\u003c/span\u003e\u003cspan address=\"10.1093/cercor/bhx179\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThomas Yeo BT, Krienen FM, Sepulcre J, et al. The organization of the human cerebral cortex estimated by intrinsic functional connectivity. \u003cem\u003eJ Neurophysiol\u003c/em\u003e. 2011;106(3):1125\u0026ndash;1165. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1152/jn.00338.2011\u003c/span\u003e\u003cspan address=\"10.1152/jn.00338.2011\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDonoho D, Jin J. Higher criticism for detecting sparse heterogeneous mixtures. \u003cem\u003eAnn Stat\u003c/em\u003e. 2004;32(3):962\u0026ndash;994. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1214/009053604000000265\u003c/span\u003e\u003cspan address=\"10.1214/009053604000000265\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDonoho D, Jin J. Higher Criticism for Large-Scale Inference, Especially for Rare and Weak Effects. \u003cem\u003eStat Sci\u003c/em\u003e. 2015;30(1):1\u0026ndash;25. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1214/14-STS506\u003c/span\u003e\u003cspan address=\"10.1214/14-STS506\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGerlach AR, Karim HT, Kazan J, Aizenstein HJ, Krafty RT, Andreescu C. Networks of worry-towards a connectivity-based signature of late-life worry using higher criticism. \u003cem\u003eTransl Psychiatry\u003c/em\u003e. 2021;11(1):550. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41398-021-01648-5\u003c/span\u003e\u003cspan address=\"10.1038/s41398-021-01648-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWilson JD, Gerlach AR, Karim HT, Aizenstein HJ, Andreescu C. Sex matters: acute functional connectivity changes as markers of remission in late-life depression differ by sex. \u003cem\u003eMol Psychiatry\u003c/em\u003e. Published online July 6, 2023. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41380-023-02158-0\u003c/span\u003e\u003cspan address=\"10.1038/s41380-023-02158-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSundermann B, Feldmann R, Mathys C, et al. Functional connectivity of cognition-related brain networks in adults with fetal alcohol syndrome. \u003cem\u003eBMC Med\u003c/em\u003e. 2023;21(1):496. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s12916-023-03208-8\u003c/span\u003e\u003cspan address=\"10.1186/s12916-023-03208-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-5005391/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5005391/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eIn this study, we aim to identify neurobiological factors that predict relapse risk in late-life depression (LLD). We recruited 145 older adults (age\u0026thinsp;\u0026ge;\u0026thinsp;60): 102 recently remitted LLD participants and 43 healthy comparisons. Participants underwent baseline MRI and evaluation of depression symptoms/status for up to 2 years. We evaluated intrinsic network connectivity for 111 participants (39 healthy comparisons, 47 stable remitted, 25 relapsed). LLD participants had lower connectivity primarily within and between the default mode (DMN), somatomotor, and visual networks and higher connectivity between the DMN and executive control network. Lower connectivity of DMN to somatomotor and salience networks was associated with relapse. Notably, the network structure of relapsed participants was more similar to healthy comparisons than stable remitted. These findings indicate that remission is associated with persistent functional network alterations while vulnerability to relapse may be associated with a failure to establish a new stable homeostatic functional network structure.\u003c/p\u003e","manuscriptTitle":"Network homeostasis: functional brain network alterations and relapse in remitted late-life depression","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-10-21 10:40:26","doi":"10.21203/rs.3.rs-5005391/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"5b28aeca-e10d-4eb8-897e-cf543f285dbd","owner":[],"postedDate":"October 21st, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":39124998,"name":"Health sciences/Diseases/Psychiatric disorders/Depression"},{"id":39124999,"name":"Biological sciences/Neuroscience/Emotion"}],"tags":[],"updatedAt":"2024-12-09T10:05:39+00:00","versionOfRecord":[],"versionCreatedAt":"2024-10-21 10:40:26","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-5005391","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5005391","identity":"rs-5005391","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}