SEARCH · PubMed Health

Results for “Memory, Short-Term”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 19 recordsLinked to original sources

PubMed2026

Diving Deeper Into Mechanisms of Acrylamide-Induced Toxicity: RNA Sequencing Reveals Transcriptomic Alteration and Retrotransposon Expression in Drosophila melanogaster.

Given the inevitability of human and animal exposure to acrylamide, there is increasing concern regarding its potential health risks. While a number of molecular mechanisms have been proposed, the complexity of acrylamide toxicological pathways and interactions remains incompletely characterized. In this study, we employed a transcriptomic approach to investigate the transcriptional responses of Drosophila melanogaster following exposure to acrylamide (100 mg/kg). Our analysis identified 634 differentially expressed genes (DEGs), with 362 upregulated and 272 downregulated. Functional analysis revealed these DEGs are enriched in pathways related to reproduction, detoxification, cellular and metabolic processes, signaling, synaptic formation and organization. Notably, acrylamide exposure upregulated the expression of tau and beta-amyloid protein precursor-like genes, both implicated in Alzheimer's disease pathology. An aversive memory test further demonstrated that acrylamide impaired the short-term memory of treated flies. Additionally, acrylamide-induced toxicity altered the expression of nine long terminal repeat retrotransposons, belonging to the gypsy and pao superfamilies. By exploring the potential role of transposable element activity in acrylamide-mediated toxicity, this study provides novel insights into the molecular mechanisms underlying its effects. Collectively, these findings offer a more comprehensive understanding of the mechanisms and pathways associated with the toxic action and detoxification of acrylamide in D. melanogaster.

Animals↗

PubMed2026

Whole-Genome Deep Learning Predicts Chemotherapy Response in Colorectal Cancer.

Chemotherapy response in colorectal cancer (CRC) exhibits significant heterogeneity, with current clinical predictors failing to capture complex genomic determinants of resistance. We developed a hybrid deep learning framework integrating convolutional neural networks (CNNs) and bidirectional long short-term memory (BiLSTM) networks to analyze whole-genome somatic mutations, evolutionary conservation, chromatin accessibility, and 3D genome architecture in 2,546 TCGA patients. An attention mechanism identified predictive genomic regions. The model achieved an AUC of 0.92 (95% CI: 0.89-0.94) in cross-validation and 0.88 (95% CI: 0.85-0.91) in independent validation, outperforming clinical models (ΔAUC = +0.18, p < 0.001). Key predictors included non-coding variants in TP53, KRAS, and PIK3CA regulatory regions. Triple-positive patients (mutations in all 3 regions) had significantly worse progression-free survival (HR = 4.7, p < 0.001). Our framework enables accurate chemotherapy response prediction and reveals novel non-coding resistance mechanisms, advancing precision oncology in CRC.

Humans↗

PubMed2026

Non-destructive prediction of lead content in oilseed rape leaves by fluorescence hyperspectral technology based on neural network.

Based on fluorescence hyperspectral imaging (FHSI), this study targeted rapid, non-destructive quantification of lead (Pb) content in oilseed rape leaves treated with varying silicon (Si) concentrations, acquiring fluorescence spectra over the 484.43-1001.61 nm wavelength range. To optimize spectral data quality, preprocessing methods (Savitzky-Golay smoothing, first derivative, detrending) were comprehensively compared. Characteristic wavelengths were then selected via interval variable iterative shrinkage, which effectively compressed data dimensionality and reduced computational load. A hybrid SE-CL1DA model, fusing a 1D convolutional neural network, a long short-term memory network and SE attention mechanism was constructed, with Bayesian optimization tuning hyperparameters to boost stability. The BO-SE-CL1DA outperformed both traditional machine learning and insufficiently optimized deep learning model (Rp2=0.9609, RMSE = 0.0377 mg/kg, RPD = 5.1736), thus enabling accurate Pb estimation, supporting Si-regulated heavy metal stress management and facilitating agricultural contamination monitoring.

Plant Leaves↗

PubMed2025

Soffritto: a deep learning model for predicting high-resolution replication timing.

MOTIVATION: Replication timing (RT) refers to the order in which DNA loci are replicated during S phase. RT is cell-type specific and implicated in cellular processes including transcription, differentiation, and disease. RT is typically quantified genome-wide using two-fraction assays (e.g. Repli-Seq) which sort cells into early and late S phase fractions followed by DNA sequencing, yielding a ratio as the RT signal. While two-fraction RT data are widely available in multiple cell lines, it is limited in its ability to capture high-resolution RT features. To address this, high-resolution Repli-Seq, which quantifies RT across 16 fractions, was developed, but it is costly and technically challenging with very limited data generated to date. RESULTS: Here, we developed Soffritto, a deep learning model that predicts high-resolution RT data using two-fraction RT data, histone ChIP-seq data, GC content, and gene density as input. Soffritto is composed of a Long Short-Term Memory (LSTM) module and a prediction module. The LSTM module learns long- and short-range interactions between genomic bins, while the prediction module is composed of a fully connected layer that outputs a 16-fraction probability vector for each bin using the LSTM module's embeddings as input. By performing both within cell line and cross-cell line training and testing for five human and mouse cell lines, we show that Soffritto is able to capture experimental 16-fraction RT signals with high accuracy, and the predicted signals allow detection of high-resolution RT patterns. AVAILABILITY AND IMPLEMENTATION: Soffritto is available at https://github.com/ay-lab/Soffritto.

Deep Learning↗

PubMed2026

SurvGRN: a multi-feature fusion framework for bladder cancer survival prediction.

Bladder cancer survival outcomes exhibit significant heterogeneity, influenced by multifaceted factors. While digital pathology-based survival models leveraging artificial intelligence show promise, they often overlook complementary data sources. Conversely, imaging lacks cellular detail, and genomics/proteomics entail complexity and cost. To integrate multidimensional data for enhanced survival prediction, we propose SurvGRN, a multi-feature fusion framework. SurvGRN synergistically combines clinical variables, transcriptomics, and digital pathology slides using a gated residual network architecture. Pathological features are extracted via multiple instance learning, while clinical and transcriptomic data are processed as static inputs. These features are dynamically fused using a long short-term memory (LSTM) network for comprehensive survival risk assessment. Evaluated on 400 bladder cancer patients, SurvGRN significantly outperformed existing methods: improving the C-index by 12.6% over DeepMISL; 20.6% and 7.1% over graph-based models (DeepGraphConv and Patch-GCN); and 5.4% and 4.0% over attention-based approaches (Surformer and HVTSurv). Ablation studies confirmed the contributions of pathology features (extracted via ResNet-50 pre-trained on bladder tissue), clinical/transcriptomic data, and the LSTM fusion. SurvGRN also enabled significant stratification of patients into distinct risk cohorts. This work demonstrates that holistic integration of multi-source data through tailored fusion architectures substantially improves bladder cancer survival prediction.

bladder cancer↗

PubMed2026

[Applications and Challenges of Deep Learning in Human Genome Research].

In recent years, the advent of high-throughput omics technologies has fueled an explosive growth in human genomic data. Uncovering the latent functions within this vast data has become a significant challenge in functional genomics research. While traditional statistical methods have proved successful for analyzing smaller-scale datasets in the past, they exhibit clear limitations in analytical efficiency and integrating multi-dimensional data, struggling to meet the escalating demands of contemporary genomic analysis. The introduction of deep learning (DL) technologies offers a novel paradigm for this field. This review systematically examines the advances in applying deep learning to human genomics research. Studies demonstrate that when ample labeled data is available, discriminative DL computational methods-such as Convolutional Neural Networks (CNNs) and Long Short-Term Memory networks (LSTMs)-achieve high accuracy and efficiency in genomic variant discovery tasks. Furthermore, generative DL methods, particularly Large Language Models (LLMs) leveraging self-supervised pre-training strategies, effectively integrate complex genomic information and exhibit superior performance in functional genomic sequence annotation and gene regulation studies. This review also explores the application of LLMs in multi-omics data integration and prediction. Looking ahead, the continued accumulation of long-read sequencing and high-dimensional data is expected to enable DL technologies to integrate increasingly complex and heterogeneous genomic information, playing an increasingly crucial role in human genomics research.

Deep Learning↗

PubMed2026

The asymmetry of working memory training transfer: A systematic review and meta-analysis.

Working memory (WM) training is widely used to enhance cognitive performance; however, its transfer to untrained tasks remains controversial. Traditional theories emphasize task similarity as the primary determinant of training transfer, but they cannot fully explain emerging evidence of asymmetric transfer across tasks. Two directional transfer hypotheses are proposed here to explain this asymmetry: the resource-based transfer advantage hypothesis predicts stronger transfer from more to less cognitively demanding tasks, whereas the ability-based transfer advantage hypothesis predicts stronger transfer from tasks engaging broader task-general abilities to tasks engaging task-specific narrower abilities. The contrast between span and updating paradigms provides an informative framework for distinguishing these accounts, because updating tasks are generally more cognitively demanding, whereas span tasks involve broader abilities. Accordingly, we conducted a three-level meta-analysis of 55 studies (208 effect sizes; N = 3,492). The results showed that updating training transferred reliably to span tasks (g = 0.176, p < .001), whereas span training did not reliably transfer to updating tasks (g = 0.048, p = .453), supporting the resource-based account. This advantage of updating training also extended to non-WM outcomes and was more pronounced at lower training doses, in non-adult samples, and with verbal stimuli. Together, these findings extend WM transfer theory beyond task similarity by highlighting the importance of cognitive demand and offer guidance for WM training design.

Humans↗

PubMed2026

Dissociable neural mechanisms of cognitive enhancement through transcranial stimulation and behavioral training.

BACKGROUND: Transcranial direct current stimulation (tDCS) and adaptive working memory (WM) training are promising cognitive enhancement approaches; however, their neural mechanisms and potential synergies remain poorly understood. OBJECTIVE: We directly compared how tDCS and WM training modulate neural oscillations during WM performance and examined whether combining both interventions produces additive effects. METHODS: We randomized 112 healthy adults into four groups: control (sham tDCS + non-adaptive 1-back), tDCS-only (active tDCS + non-adaptive 1-back), training-only (sham tDCS + adaptive n-back training), or combined (active tDCS + adaptive training). Participants underwent five daily intervention sessions. We recorded high-density EEG during transfer n-back tasks at baseline, post-intervention, and one-week follow-up. RESULTS: All active interventions improved WM performance relative to the control group, with the combined group showing the largest gains (n-back accuracy: +15.6% vs. + 10.1% tDCS-only, +9.7% training-only, +0.7% control; all p < 0.001). Critically, tDCS selectively increased gamma-band (30-50 Hz) power in the frontal and parietal regions (cluster p = 0.018, d > 1.0), whereas WM training enhanced frontal theta-band (4-8 Hz) power and theta-gamma phase-amplitude coupling (both cluster p < 0.012, d > 0.85). The combined group exhibited both neural signatures. Brain-behavior correlations revealed dissociable relationships: gamma increases predicted n-back accuracy improvements (r = 0.61, p < 0.001), whereas theta enhancements correlated with operation span gains (r = 0.58, p = 0.002). CONCLUSIONS: tDCS and WM training enhance cognition through distinct yet complementary neural mechanisms: tDCS via gamma-mediated cortical excitability and WM training via theta-mediated cognitive control. These findings provide neurophysiological evidence for multimodal enhancement strategies that target parallel pathways within WM networks.

Humans↗

PubMed2026

Childhood Trauma and Frontoparietal Network Connectivity During Working Memory Task Performance in Individuals With Schizophrenia and Healthy Participants.

Schizophrenia is associated with altered frontoparietal connectivity, which supports higher-order cognition, including working memory. Childhood trauma has been linked to altered functional connectivity and reduced cognitive performance in individuals with schizophrenia and controls. Prior evidence suggests that trauma-related default mode dysconnectivity mediates the association between trauma and cognition. We hypothesised that childhood trauma would be associated with altered frontoparietal connectivity during a working memory task and that such connectivity changes would mediate the relationship between trauma and working memory. Childhood trauma, working memory and fMRI data were collected from individuals with schizophrenia or schizoaffective disorder (n = 38) and controls (n = 128). fMRI data were preprocessed in CONN, and seed-based connectivity analyses were performed for four frontoparietal seeds (bilateral dorsolateral prefrontal and posterior parietal cortices). Connectivity was contrasted across (a) diagnosis and (b) trauma severity. Moderated mediation analyses tested the associations between trauma, connectivity and working memory, with diagnosis as a moderator. Across all participants, higher physical neglect severity predicted poorer working memory performance. Stronger inverse connectivity between left dorsolateral prefrontal and frontal medial cortices predicted better working memory performance. As expected, patients showed widespread frontoparietal dysconnectivity relative to controls, but no differences in frontoparietal connectivity were observed across trauma severity groups. Frontoparietal connectivity did not mediate the association between trauma and working memory, although diagnosis moderated both the trauma-connectivity and connectivity-cognition associations. We conclude that, unlike previous evidence suggesting a mediating role for the default mode network, frontoparietal connectivity did not mediate the trauma-cognition association, possibly suggesting the unique significance of default mode network dysconnectivity in linking trauma to cognition in psychosis.

Humans↗

PubMed2026

Effects of quetiapine on cognitive functioning in schizophrenia: evidence for the remyelination hypothesis?

Postmortem findings, neuroimaging data, and in-vitro models suggest a decrease in number and density of oligodendrocytes is driving cognitive deficits in schizophrenia (SCZ). Second-generation antipsychotics are discussed to improve oligodendrocyte dysfunction with most conclusive evidence available for quetiapine (QET). We postulate that sustained QET treatment leads to cognitive improvement in SCZ, particularly, in tests with high demands for working memory function. We further hypothesize that these effects are moderated by polygenic factors associated with hippocampus-related brain volumes, general white matter integrity, and/or oligodendroglia-related SCZ risk. Using data of the prospective PsyCourse study, we identified 166 patients with SCZ spectrum disorder receiving QET at one or two consecutive visits plus 166 matched patients without QET. Polygenic scores were calculated for subcortical brain volumes, measures of white matter integrity, and for cell type-specific genetic SCZ risks. QET treatment was consistently associated with improved cognitive function independent of time, specifically, in tests with high, but not with low to medium working memory load. Polygenic analyses did not reveal significant moderation effects. In contrary, low genetic SCZ risk specific for genes related to human oligodendrocyte function was associated with higher cognitive performance independent from QET. While we observed improved cognitive performance under QET in high working memory tests, we did not find evidence that polygenic factors associated with hippocampus-related brain volumes, white matter integrity, or oligodendroglia-related SCZ risk moderate this association. Thus, our tentative findings do not provide evidence for the hypothesis that polygenic estimates of hippocampal remyelination capacities influence the association between QET and cognitive performance in SCZ.

Humans↗

PubMed1975

Effect of two weeks' treatment with thioridazine, chlorpromazine, sulpiride and bromazepam, alone or in combination with alcohol, on learning and memory in man.

Forty paid healthy male students participated in two subacute experiments of 6 weeks each. In the first trial 20 of them received bromazepam, thioridazine, and placebo double blind cross over for 2 weeks each, and in the second trial the active agents administered to the other 20 participants were chlorpromazine and sulpiride. The tests used were paired associate learning with nonsense syllables and digit memory span. Before testing the subjects took either an alcoholic or a nonalcoholic bitter drink. As in the previous study from this laboratory, alcohol was found to impair learning capacity. Of the drugs used only bromazepam impaired learning significantly, and the combined effect of alcohol and bromazepam on learning capacity was very deleterious. The adrenolytic effect of drugs did not correlate with their effect on learning. Caution is necessary when prescribing bromazepam for active outpatients at least in doses used in this study.

Adult↗

PubMed2025

A comparative study highlights superiority of LSTM in crop genomic prediction.

We systematically evaluated three key determinants affecting prediction accuracy and the algorithm performance differences based on fifteen state-of-the-art GP methods, and found LSTM suitable for capturing additive and epistatic effects. Genomic prediction (GP) has been developed as an important method supporting crop breeding. By utilizing the phenotype values result from GP, breeders could make decisions in the seedling stage that consequently benefit for cost saving. In recent years, machine learning emerged as an efficient technology to solve modeling problems in many fields, including crop breeding. However, numerous modeling approaches have hindered the application of GP since breeders struggle to choose. Therefore, a comprehensively methodological research with guiding significance is extremely necessary. In the present study, we systematically evaluated three key determinants affecting prediction accuracy and the algorithm performance differences based on fifteen state-of-the-art GP methods. As for genomic feature processing, we found feature selection (SNP filtering approach) performed better than feature extraction (PCA method). Specifically, the feature relationship dependent methods (GBLUP, RNN, and LSTM) as well as DNN architecture showed superior performance with feature selection. Marker density analysis showed positive correlation with prediction accuracy in a limited threshold. Comparison on effect of population size demonstrated a positive correlation between trait genetic complexity and the optimal population size required. By testing fifteen modeling methods, we found LSTM network displayed superior performance, achieving the highest average STScore (0.967) across six datasets. Further research using all cell states or the latest cell states of LSTM inputs demonstrated its architecture particularly adept with capturing additive and epistatic QTL effects among SNPs. In conclusion, our findings provide basic principles for implementing GP in breeding project to maximize prediction accuracy while maintaining cost-effectiveness.

Plant Breeding↗

PubMed2026

Effectiveness of transcranial direct current stimulation with and without positive mood induction on worry and transdiagnostic cognitive-emotional processes: A randomized controlled trial.

The present study investigated the effectiveness of transcranial direct current stimulation (tDCS), with and without positive mood induction, on worry and key transdiagnostic cognitive-emotional processes, including attentional bias, working memory, problem solving, and emotion regulation, in individuals with high levels of worry. This single-blind randomized controlled trial included 45 individuals with high levels of worry. After a structured clinical interview, participants were randomly assigned, with gender balancing, to one of three groups: (1) tDCS alone, (2) tDCS combined with positive mood induction, or (3) a sham control group. Outcome measures were administered at three time points (pretest, posttest, and one-month follow-up) and assessed attentional bias (Dot Probe Task), working memory (1-back task), problem solving (Tower of London task), emotion regulation (Gross's Emotion Regulation Questionnaire), and worry severity (Penn State Worry Questionnaire; PSWQ). Repeated-measures ANOVA showed that both active groups (tDCS alone and tDCS + positive mood induction) significantly improved attentional bias, worry, working memory, problem solving, and emotion regulation compared to controls (p < 0.05). The combined intervention produced significantly greater gains than tDCS alone in working memory, problem solving, emotion regulation (p < 0.05), and reductions in attentional bias and worry (p < 0.001). All effects persisted at one-month follow-up (p < 0.05). tDCS reduces worry and attentional bias and enhances cognition and emotion regulation in individuals with high levels of worry. The combined intervention produced larger and more sustained improvements than tDCS alone across the assessed behavioral outcomes. These findings support further investigation of combining tDCS with structured positive mood induction while the mechanisms underlying the additional benefits remain to be established.

Humans↗

PubMed2025

Genetic and epigenetic changes to the glucocorticoid receptor gene (NR3C1) and cognition in major depressive disorder.

INTRODUCTION: Many studies have found that hypothalamic-pituitary-adrenal (HPA) axis abnormalities are related to the pathophysiology of major depressive disorder (MDD) and cognitive functioning. Our aim was to assess the influence of genetic polymorphisms and methylation levels in three different promoter regions throughout the glucocorticoid receptor (GR) gene NR3C1 on cognitive performance in MDD. Plausible interactions with childhood adversity and mediation relationships between genetic and epigenetic variables were explored. MATERIALS AND METHODS: The sample included a total of 64 MDD patients and 82 healthy controls. Child maltreatment and neurocognitive performance were assessed in all participants. HPA negative feedback was analyzed using the dexamethasone suppression test after the administration of 0.25mg of dexamethasone. A total of 23 single-nucleotide polymorphisms were genotyped, and methylation levels at several CpGs in exons 1D, 1F and 1H of the GR gene were measured. RESULTS: Results show that, beyond the influence of other covariables, NR3C1 single-nucleotide polymorphisms and methylation levels predicted performance in executive functioning and working memory tasks. No significant interactions or mediation relationships were detected. CONCLUSIONS: Results suggest that genetic variations and epigenetic regulation of the GR gene are relevant factors influencing cognitive performance in MDD and could emerge as significant biomarkers and therapeutic targets in mood disorders and other stress-related disorders.

Humans↗

PubMed2026

Distinct contributions of schizophrenia and neurotransmitter pathway genetic liability to neurocognition and antipsychotic efficacy in drug-naïve first-episode schizophrenia.

The genetic mechanisms underlying heterogeneity in symptom presentation and antipsychotic response in schizophrenia remain unclear, limiting the development of personalized treatment. We integrated genome-wide schizophrenia polygenic risk scores (SZ-PRS) and pathway-specific PRSs (pPRSs) for four major neurotransmitter systems to examine their associations with clinical phenotypes across the course of illness. Primary analyses were conducted in 394 drug-naïve, first-episode patients from the Chinese First-Episode Schizophrenia Trial (CNFEST) to investigate associations with baseline symptom severity, neurocognitive impairment, and longitudinal treatment response. The CNFEST cohort included 52-week longitudinal assessments of symptoms and neurocognition using the Positive and Negative Syndrome Scale and a modified version of the MATRICS Consensus Cognitive Battery. An independent case-control cohort evaluated associations with schizophrenia diagnosis, while a cohort of 514 healthy adults assessed whether PRS-cognition associations are specific to schizophrenia. Higher SZ-PRS predicted schizophrenia diagnosis (OR = 2.28, Pfdr = 0.003) and poorer baseline executive function (β = -0.44, Pfdr = 0.006) and working memory (β = -0.49, Pfdr = 0.018), but these associations were absent in healthy adults. In contrast, pPRSs showed weaker associations with diagnosis and baseline cognition but were more informative for treatment outcomes: higher serotonin-pPRS predicted greater improvement in depressive symptoms (Pfdr = 0.023-0.032), and higher GABA-pPRS predicted greater improvement in overall symptoms (Pfdr = 0.038-0.043) during weeks 4-24. Exploratory drug-specific analyses further suggested that treatment response varied across antipsychotics and was differentially associated with pPRSs. These findings demonstrate that genome-wide and pathway-specific PRSs contribute distinctly to schizophrenia phenotypes, supporting their integration for personalized stratification and treatment.

Humans↗

PubMed2026

Electromagnetic Radiation Stimulated Learning in Perovskite Nickelates.

Biological plasticity refers to the ability of synapses to strengthen or weaken over time. These adaptive properties play a fundamental role in learning and memory, spanning many orders of magnitude in timescales. Short-term plasticity (STP) arises from rapid correlative activity, while long-term plasticity (LTP) is governed by slower biochemical processes. Here, we investigate electromagnetically driven relaxation dynamics in perovskite nickelate thin films as an analogue of biological learning behaviors. By comparing radio frequency (RF), infrared (IR), visible, and ultraviolet (UV) radiation as stimuli, we find that RF excitation primarily induces STP, while visible and IR illumination lead to reversible relaxation on behavioral timescales. In contrast, UV illumination results in persistent, non-thermal changes in conductivity over extended timescales. Notably, UV-exposed nickelate films exhibit glass-like dynamics, characterized by stretched exponential relaxation and aging phenomena. The films display habituation to repeated stimuli, along with sensitization and spontaneous recovery under controlled environments. A minimal dynamical systems model captures key qualitative features of the UV-induced resistance changes. Our results demonstrate that electromagnetic frequency enables multi-timescale relaxation spanning nearly nine orders of magnitude, suggesting perovskite nickelates as promising platforms for adaptive optoelectronic hardware and for linking computational neuroscience with emerging quantum technologies.

electromagnetic radiation↗

PubMed2025

Impact of Chewing Behavior Change on Cognition and Cerebral Hemodynamics.

BACKGROUND: Impaired chewing ability is a recognized risk factor for cognitive decline in older adults, potentially due to reduced neural stimulation in cognition-related brain regions. While short-term studies have demonstrated transient increases in neural activity from chewing, the sustained cognitive and neurophysiological effects of encouraging thorough chewing habits in daily life remain unclear. OBJECTIVE: This randomized controlled trial investigated whether promoting thorough chewing during meals could improve cognitive function and cerebral hemodynamics in older adults. METHODS: Fifty participants aged 65 y or older were randomly assigned to either a 1-mo intervention group, which used a wearable device to monitor and increase chewing strokes during meals, or a control group that maintained usual chewing habits. Chewing behavior, cognitive performance (including memory and executive function via the color Stroop test), and cerebral hemodynamics in the dorsolateral prefrontal cortex (DLPFC) were measured at baseline and after 1 mo. Statistical analyses included t tests, chi-square tests, 2-way analysis of variance with post hoc tests, Pearson correlations, and generalized linear models to evaluate group differences and associations between chewing and cognitive outcomes. RESULTS: Significant time-by-group interactions were observed for memory, F(1, 48) = 6.24, P = 0.043, and hemodynamic responses in the left DLPFC, F(1, 48) = 6.19, P = 0.013. The intervention group showed increased chewing frequency (P = 0.017), improved memory performance, and reduced left DLPFC responses compared with controls. Chewing frequency was positively correlated with Stroop test scores (r = 0.53, P = 0.010) and negatively with hemodynamic changes in the left DLPFC (r = -0.30, P = 0.040). Although improvements in other cognitive outcomes and hemodynamic measures favored the intervention group, these differences did not reach statistical significance. CONCLUSIONS: Promoting intentional chewing habits for 1 mo may enhance memory-related cognitive performance and neural efficiency in the DLPFC during working memory tasks in older adults. This nonpharmacologic, low-burden strategy warrants further research with longer interventions to support cognitive health and dementia prevention. TRIAL REGISTRATION ID: UMIN000044280Knowledge Transfer Statement:This study demonstrates that promoting thorough chewing habits in older adults can improve memory and enhance neural efficiency in the brain. Encouraging intentional mastication is a simple, nonpharmacologic approach that may help maintain cognitive health and prevent dementia, providing a practical strategy for clinicians and policymakers to support healthy aging.

Humans↗

PubMed2026

Comparing trajectories of cognitive functioning in treatment-resistant and non-resistant depression: a multicentre linear mixed-effects analysis.

BACKGROUND: Impaired cognitive functioning is a severe symptom in major depressive disorder (MDD). Recent evidence suggests it may be a central characteristic in its treatment resistant form (TRD), potentially constituting a clinical marker for treatment resistance and a target amenable to intervention. To date, cognitive functioning in TRD remains poorly understood and longitudinal investigations are scarce. METHODS: This observational prospective cohort study, including 320 patients diagnosed with MDD from the multicentre PROMPT study, examined differences in cognitive functioning between 118 TRD and 202 non-TRD patients over a period of twelve weeks in a real-world setting, using linear mixed modelling. Patients that failed to respond to at least two prior antidepressants trials at baseline were classified as TRD. RESULTS: TRD patients showed significantly poorer baseline performances than non-TRD patients in attention/processing speed (β = -0.45; 95%CI[-0.70, -0.19]; FDR-p = 0.003) and verbal memory (β = -0.45; 95%CI[-0.72, -0.18]; FDR-p = 0.003). Significant time × group interactions were observed in motor speed and verbal fluency tasks. Post-hoc-analyses revealed stagnation in TRD patients and significant improvement in non-TRD patients. Across all other tasks improvement was observed in both groups, and random effects showed large heterogeneity between patients, indicating notable individual differences in cognitive performances. CONCLUSIONS: The results suggest distinct recovery patters between non-TRD and TRD patients, and diminished functioning in TRD patients at the domain level. However, intact and diminished performances likely occur in both groups, warranting further investigation of cognitive heterogeneity. These short-term findings highlight the need for more comprehensive longitudinal research on cognition in TRD.

Humans↗