The hippocampus is central to Alzheimer’s disease (AD), characterized by atrophy and cognitive decline. While volume loss is well-documented, surface-based morphometric (SBM) features—curvature, gyrification, and thickness—remain less explored. Using T1-weighted MRI data from the Alzheimer’s Disease Neuroimaging Initiative (3,144 timepoints; CN: 450, MCI: 284, AD: 204), hippocampal subfields were analyzed with HippUnfold. Linear mixed effects models examined volume and SBM changes, tracking cognitive trajectories in stable (n = 710) and progressing (n = 228) individuals.Focusing on CA1 as a representative subfield, baseline volume differed significantly across diagnostic groups, with all clinical groups showing reduced volume relative to cognitively normal individuals, consistent with disease severity. In contrast, surface-based morphometry (SBM) measures showed minimal baseline group differences, with only the CN to MCI/AD group exhibiting a significant negative main effect across SBM metrics. Longitudinal analyses revealed that disease-related changes were primarily captured by time-dependent interactions. In particular, the MCI to AD group demonstrated the most robust longitudinal changes, showing the largest magnitude of volume loss and accompanying SBM alterations over follow-up. Individuals progressing from CN to MCI/AD exhibited similar, though slightly smaller, longitudinal SBM changes, indicating early surface remodeling prior to overt clinical progression. In contrast, Stable AD showed significant longitudinal effects only for volume loss, and Stable MCI displayed minimal or nonsignificant SBM and volumetric changes. Analyses linking longitudinal morphometric change to cognitive domains was differentially associated with rates of cognitive decline across domains and disease stages, with SBM changes linked to slower language decline in early stages and to accelerated memory and visuospatial decline during progression from MCI to AD.While hippocampal volume loss remains a robust marker of AD, it does not fully capture longitudinal morphometric changes associated with disease progression. In particular, SBM measures showed their strongest and most consistent longitudinal effects in individuals progressing from MCI to AD, where changes in surface geometry accompanied accelerated volume loss. These findings suggest that SBM features are most sensitive to morphometric reorganization during active disease progression and provide complementary information beyond volume alone.

OBJECTIVE:To identify which patients with remitted major depressive disorder (rMDD) or mild cognitive impairment (MCI) benefit from cognitive remediation (CR) plus transcranial direct current stimulation (tDCS). DESIGN:We conducted a moderator analysis to examine the effects of baseline brain magnetic resonance imaging (MRI) measures on the impact of CR + tDCS on cognitive decline in Prevention of Alzheimer's dementia with CR plus tDCS in MCI and Depression (PACt-MD), a double-masked randomized two-arm controlled trial with assessments at baseline, two months, and yearly for three to seven years. SETTING:Five academic hospitals in Toronto, Canada. PARTICIPANTS:A total of 246 participants with rMDD, MCI, or both, with an analyzable baseline MRI. INTERVENTION:CR + tDCS or sham CR + sham tDCS. MEASUREMENTS:Overall cortical thickness, overall fractional anisotropy, and cortical thickness in an a-priori composite region of interest (ROI); changes in global cognition, executive function, or verbal memory. RESULTS:Overall cortical thickness moderated decline in global cognition (Χ² = 10.43, df = 3, p = 0.015); ROI cortical thickness moderated treatment-related changes in global cognition (Χ² = 29.05, df = 3, p <0.001), executive function (Χ² = 11.57, df = 3, p = 0.009), and verbal memory (Χ² = 16.08, df = 3, p = 0.001). CONCLUSION:Future work needs to confirm that cortical thickness can be used to select adults at risk for dementia who are the most likely to benefit from CR + tDCS. CLINCIALTRIALS. GOV IDENTIFIER:NCT02386670.

Cognitively Unimpaired (CU), Mild Cognitive Impairment (MCI), and Alzheimer's Disease (AD) are clinical labels used to categorize degrees of cognitive impairment in the aging brain. Older adults experience brain changes associated with aging and cognitive decline at different ages and progress at varying rates, leading to heterogeneous patterns of cognitive decline. As a result, people within the same diagnostic category often exhibit significant differences in cognitive abilities and brain functions. The goal of the present study was to investigate how data-driven categories map onto diagnostic categories. We combined brain imaging features (cortical thickness average, surface area, and volume) with cognitive measures (Alzheimer's Disease Assessment Scale-Cognitive, Mini-Mental State Exam, Montreal Cognitive Assessment, Clinical Dementia Rating) in older adults who had a clinical diagnosis of CU, MCI, or AD using Similarity Network Fusion (SNF), a multivariate clustering approach. SNF is a new computational method for data integration that leverages common and complementary information in different types of data. We used data from 515 participants in the Alzheimer's Disease Neuroimaging Initiative (ADNI) database. We identified four data-driven groups spanning a gradient of cognitive and neural severity, with an average silhouette width of 0.55, indicating good cluster structure. Group 1 (87% diagnosed with dementia) showed the greatest impairment, while Group 4 (96% cognitively unimpaired) showed minimal impairment. Groups 2 and 3 captured transitional stages, including an “at-risk” group with early neural and cognitive decline. The top contributing features included the right lingual surface area (NMI = 0.34) and MMSE scores (NMI = 0.052). Significant demographic differences were observed across clusters (e.g., age: F(3, 511) = 30.39, p  < 0.001). The current study provides evidence that more nuanced, data-driven approaches can reveal commonalities in the etiology and underlying neurobiology of individuals across traditional categories of CU, MCI, or AD. These results may lead to more focal therapies and a better understanding of who is at risk for converting to dementia, allowing for earlier detection and treatment of cognitive decline.

Treatment for major depressive disorder (MDD) remains challenging as only 30-50% of patients respond to first-line antidepressant medications in primary care. Here we developed algorithms using predictors of response to sertraline and bupropion from a multisite study, and tested such markers in an independent, prospective clinical trial involving unmedicated individuals with MDD (NCT05537584). Leave-one-out cross-validation models achieved good performance in the training sample (area under the curve of 0.66-0.86). In the preregistered clinical trial, no significant differences in treatment outcomes emerged for those assigned a drug consistent versus inconsistent with their biomarkers. However, significant differences emerged in symptom reduction trajectories for those with positive markers for both medications (response rate: 71.4%) or either drug (65.4%) compared with those with two negative markers (42.9%). This is the first study using biobehavioral markers to prospectively guide assignment to two widely used antidepressants, yielding a 66.8% boost in response rate and providing foundations for larger personalized treatment studies of MDD.

OBJECTIVE:Everyday functional capacity in older adults is influenced by several factors, with prior studies finding that cognition mediates the relationship between depression and everyday functioning. However, these studies utilized samples with low depression severity and used only one type of functional assessment. We aimed to examine whether cognition mediates the relationship between depression and functioning in older adults with a history of treatment-resistant depression. METHOD:Data from 383 participants enrolled in the OPTIMUM Neuro study were analyzed. Participants completed a neuropsychological assessment battery, depression severity interview, self-/informant-rated functioning measures and a performance-based functioning measure. Linear regression was used to determine whether depression scores predicted cognitive domain and everyday functioning scores. Cognitive domain scores predicted by depression were then tested as mediators between depression and functioning. RESULTS:Higher depression symptoms predicted poorer performance on all measures of functioning as well as the cognitive domains of attention, executive functioning, and immediate memory. Immediate memory partially mediated the relationship between depression and a performance-based measure of functioning, while attention and executive functioning partially mediated the relationship between a self-report measure of functioning and depression. CONCLUSIONS:The relationship between depression severity and poorer functional performance was partially mediated by attention, executive functioning, and immediate memory, with results differing based on the measure of functioning used. Our findings suggest that there may be additional non-cognitive factors influencing this relationship and highlight the importance of using multiple methods to assess functional performance.

Treatment for major depressive disorder (MDD) remains challenging as only 30% of patients respond to antidepressants in primary care. Here, we developed algorithms using predictors of response to sertraline and bupropion from a multi-site study, and tested such markers in an independent prospective clinical trial of unmedicated individuals with MDD. Leave-one-out cross-validation models achieved good performance in the training sample (AUC=0.66-0.86). In the pre-registered clinical trial, no significant differences in treatment outcomes emerged for those assigned a drug-consistent vs inconsistent with their biomarkers. However, significant differences emerged in symptom reduction trajectories for those with positive markers for both medications (response rate: 71%) or either drug (65%) compared to those with two negative markers (43%). This is the first study using biobehavioral markers to prospectively guide assignment to two widely used antidepressants. Future studies will further optimize algorithms to guide antidepressant prescription to achieve faster and stronger symptom reduction.

Historically, aperiodic neural activity (i.e., the 1/f exponent and offset of electroencephalogram (EEG) power spectra) has been treated as neural noise. However, recent work has highlighted associations among aperiodic activity, aging, cognition, and psychopathology. While depression has been associated with cognitive deficits and dysregulation of excitatory/inhibitory neural mechanisms linked to aperiodic activity, to date, few studies have compared aperiodic activity in adults with and without depression, and none have evaluated aperiodic activity as a function of depressive chronicity. Here, individuals with depression (n=72) and healthy controls (n=34) completed a resting state EEG recording from which aperiodic activity estimates were extracted. We show significant group differences in aperiodic exponent and offset across central and posterior regions, such that, as hypothesized, these parameters were lower for depressed individuals relative to controls. Follow-up analyses clarified that this effect was driven by the number of lifetime depressive episodes experienced. Future research is needed to understand how depression heterogeneity impacts aperiodic activity.

The hippocampus plays a critical role in Alzheimer's disease (AD), marked by brain atrophy and cognitive decline. AD pathology begins during the transition from healthy aging to mild cognitive impairment (MCI) and eventually AD, with memory impairment as a hallmark. While hippocampal volume reductions are well-documented, surface-based morphometric (SBM) features—curvature, gyrification, and thickness—remain less understood. T1-weighted MRI data from 3.51 average annual scans (5,263 timepoints) across four phases of the Alzheimer's Disease Neuroimaging Initiative (ADNI) were analyzed using HippUnfold, a hippocampal subfield segmentation tool. Individuals (CN: 475; MCI: 673; AD: 269) were grouped by final clinical diagnosis to track cognitive trajectories: stable/non-progressors ( n  = 1017) and progressors ( n  = 301). Linear mixed effects models adjusted for age, education, sex, scanner-site, and eTIV evaluated hippocampal volume and surface metrics (CA1–CA4, DG, Subiculum, SRLM, and Cysts), with CN as the intercept. Focusing on CA1, the region most vulnerable to early AD, the stable AD group showed significant volume reductions (β = -1.05, p  < .001) compared to CN (β = 3.88, p  < .001). All SBM metrics (curvature, gyrification, and thickness) showed significant changes, with curvature in CN (β = -0.48, p  < .001) and AD (β = 0.01, p  < .001). Time-dependent interactions showed volume reductions in CN (β = -0.04, p  < .001) and MCI (β = -0.07, p  < .001), and increases in SBM metrics (all p 's < .001). Cognitive domain analyses showed volume changes primarily affect memory and executive functioning, while SBM metrics influence language and visuospatial ability. Curvature influenced language (61.76%, p  < .001) and visuospatial ability (32.35%, p  < .001), while thickness affected language (57.14%, p  < .001) and visuospatial ability (35.71%, p  < .001). Hippocampal volume reductions are well-established markers of AD, but surface-based features like curvature, gyrification, and thickness provide additional insights, revealing changes that volume alone may miss. These findings highlight the importance of integrating surface-based metrics with volumetric analyses to improve understanding of disease mechanisms and interventions.

Background:Major depressive disorder (MDD) and treatment-resistant depression (TRD) have each been characterized by altered neural connectivity largely associated with the triple model of the default mode (DMN), frontoparietal (FPN), and salience (SN) networks. However, the direction (i.e., hyper- vs. hypoconnectivity) and the specificity (i.e., depression broadly vs. TRD) of these alterations remains unclear. Thus, in the current study, we compared high-frequency between- and within-network resting-state functional connectivity (rsFC) in healthy control (HC) individuals, individuals with MDD, and individuals with TRD. Methods:Ninety-six channel resting-state electroencephalogram data were collected from 34 participants with MDD (22 women, mean ± SD age: 29.92 ± 9.57 years), 24 participants with TRD (16 women, age: 44.35 ± 15.86 years), and 34 HC participants (25 women, age: 32.49 ± 14.07 years). Based on previous findings, exact low-resolution electromagnetic tomography was used to estimate theta and beta rsFC within and between the DMN, FPN, and SN. Results:Participants with depression (i.e., pooled MDD and TRD participants) had enhanced within-DMN beta1 (12.5-18 Hz) connectivity compared with controls. Compared with MDD participants, participants with TRD showed increased within-DMN, DMN to FPN, and FPN to SN beta3 (21.5-30 Hz) connectivity. These effects persisted when controlling for current depressive symptoms. Conclusions:Differences in high-frequency rsFC, particularly in the beta bands, among the DMN, FPN, and SN may partially account for neural mechanisms of treatment resistance. However, future work probing the heterogeneity (e.g., medication status, age of onset, lifetime episode count) and time course (e.g., length and frequency of episodes) of depression is needed to increase our understanding of these changes in neural connectivity.

Older adults with treatment-resistant depression are at significant risk for cognitive impairment. The relationship between treatment response and cognitive function in this population is not well-established. We examined neural correlates of executive and memory function, and their relationship with prospective treatment outcomes. In the context of a longitudinal biomarker study embedded within a multi-center randomized controlled trial for late-life treatment-resistant depression, 397 participants completed baseline neuropsychological testing, and of these 234 adults successfully completed a baseline MRI scan. Multivariate regressions were used to test for brain-cognition associations between memory and executive function and brain functional connectivity, white matter integrity, and gray matter structure. Further, we employed regularized elastic net regressions to identify biomarkers predicting depression remission (MADRS≤10) in the clinical trial. Among participants who completed neuroimaging better cognition was associated with lower connectivity between components of the default mode and the frontoparietal networks and within the frontoparietal network (multivariate r=0.37, p<0.01). Using diffusion imaging data, lower tract integrity in a distributed set of tracts was associated with poorer executive function (multivariate r=0.27, p<0.05). Additionally, gray matter structure was positively associated with cognition (multivariate r=0.38, p<0.05). Education and better structural brain maintenance but not overall health were associated with better cognition. Ongoing treatment resistance was predicted by poorer cognition and gray matter structure. We identified distinct cross-sectional associations between specific neural circuits and variation in cognitive function in people with treatment-resistant late-life depression. We also found worse cognitive function and gray matter structure predicted ongoing treatment resistance to medication offered in the clinical trial.