Background Amnestic mild cognitive impairment (aMCI) is considered a prodromal phase of Alzheimer’s disease (AD). However, little is known about the neuropsychological characteristic at pre-MCI stage. This study aimed to investigate which neuropsychological tests could significantly predict aMCI from a seven-year longitudinal cohort study. Methods The present study included 123 individuals with baseline cognitive normal (NC) diagnosis and a 7-year follow-up visit. All the subjects were from the China Longitudinal Aging Study (CLAS) study. Participants were divided into two groups, non-converter and converter based on whether progression to aMCI at follow-up. All participants underwent standardized comprehensive neuropsychological tests, including the mini-mental state examination (MMSE), Montreal Cognitive Assessment (MoCA), auditory verbal learning test (AVLT), the digital span test, the verbal fluency test, the visual recognition test, the WAIS picture completion task, and WAIS block design. Logistic regression analysis was used to evaluate the predictive power of baseline cognitive performance for the transformation of amnestic mild cognitive impairment. Receiver operating characteristic (ROC) curve was used to test the most sensitive test for distinguishing different groups. Results Between the non-converter group and converter group, there were significant differences in the baseline scores of AVLT-delayed recall (AVLT-DR) (8.70 ± 3.61 vs. 6.81 ± 2.96, p = 0.001) and WAIS block design (29.86 ± 7.07 vs. 26.53 ± 8.29, p = 0.041). After controlling for gender, age, and education level, converter group showed lower baseline AVLT-DR than non-converter group, while no significant difference was found in WAIS block design. Furthermore, converter group had lower AVLT-DR score after controlling for somatic disease. The area under the curve of regression equation model was 0.738 (95%CI:0.635–0.840), with a sensitivity 83.9%, specificity of 63.6%. Conclusions Our results proved the value of delayed recall of AVLT in predicting conversion to aMCI. Early and careful checking of the cognitive function among older people should be emphasized.

BackgroundAnti-amyloid disease-modifying therapies (DMTs) for early Alzheimer's disease (AD) are entering routine care, increasing the need for harmonized, registry-ready real-world data. The International Registry for Alzheimer's Disease and Other Dementias (InRAD) proposed a minimum dataset (MDS) and extended dataset (EDS), but their applicability to psychiatry-led old age mental healthcare practices in China is uncertain.ObjectiveTo adapt the InRAD dataset for real-world AD DMT practice across multiple psychiatry institutions in China and assess the feasibility of routine data capture for the proposed MDS/EDS.MethodsWe conducted a modified Delphi consensus study and a multicenter feasibility survey. Forty-nine experts classified domains/items into the MDS or EDS using predefined agreement thresholds. Thirty-five DMT-initiating mental healthcare teams reported the routine availability of the proposed data elements.ResultsHighly consistent with InRAD, ten domains were included in the China-adapted MDS/EDS, covering patient profiles and lifestyle, diagnostic work-up and biomarkers, treatment, outcomes, safety, treatment-monitoring examinations, and registry discontinuation. However, item prioritization reflected local practice, emphasizing diagnostic traceability, functional and neuropsychiatric outcomes, caregiver burden, and structured safety capture. Feasibility results revealed that many MDS elements were collected, but the consensus-defined MDS exceeded what is currently captured in a standardized, analysis-ready format; most EDS items were moderately feasible, while WHO-5 (patient version) and DAT-scan were least feasible.ConclusionsAn InRAD-aligned dataset is broadly acceptable for psychiatry-led AD DMT practices in China, but implementation gaps remain. A phased registry approach with standardized definitions and workflow-supported capture may improve the completeness and comparability of real-world DMT evidence.

Neuropsychiatric symptoms (NPS) may signal dementia risk or early pathology, yet clinical evidence regarding the association between pre-diagnostic NPS and outcomes remains limited. This preliminary study explored NPS features and their potential associations with dementia progression. The study was conducted on 201 dementia patients from Shanghai Mental Health Center (ICD-10 diagnosed). NPS history was extracted from medical records. Statistical analyses included chi-square tests (demographics/subtypes), ANOVA (preclinical intervals), and logistic regression (AD vs. non-AD factors). 69.2

[This corrects the article DOI: 10.3389/fneur.2026.1853336.].

BackgroundNeurological disorders can manifest with psychiatric symptoms as the initial presentation (PSIP), leading to diagnostic challenges and delayed treatment. Data on this patient population remain limited.MethodsWe conducted a retrospective cohort study of patients admitted to a tertiary neurology department between 2019 and 2023 with PSIP. A structured screening protocol using predefined behavioral keywords was applied, followed by independent case review by two neurologists and a psychiatrist to confirm eligibility. Final diagnoses were adjudicated by consensus using contemporary diagnostic criteria. Poor outcome was defined as modified Rankin Scale score >2 at discharge. Multivariable logistic regression was used to identify independent predictors of poor outcome, and a risk score was developed based on these factors.ResultsOf 16,473 screened admissions, 76 patients (mean age 55.8 ± 18.2 years; 42.1% female) met inclusion criteria. Most patients presented with acute-onset (92.1%) non-specific behavioral disturbance (63.2%), and nearly one-third (31.6%) had isolated psychiatric symptoms without any accompanying neurological signs. The leading etiologies were CNS infections (55.3%), predominantly viral encephalitis (40.8%), followed by cerebrovascular diseases (14.5%) and autoimmune encephalitis (13.2%). Poor outcome occurred in 32 patients (42.1%). Independent predictors of poor outcome were hyponatremia on admission (OR, 3.9; 95% CI, 1.3–11.8), viral encephalitis etiology (OR, 3.1; 95% CI, 1.1–8.7), and ICU admission (OR, 4.8; 95% CI, 1.3–17.6). A risk score combining these three factors effectively stratified patients (AUC 0.88; 95% CI, 0.80–0.96), with poor outcome rates ranging from 7.1% (score 0) to 100% (score 3). Subgroup analyses revealed that among viral encephalitis patients, those with PSIP had significantly higher rates of hyponatremia (41.9% vs. 15.2%, p = 0.002) and poorer outcomes (58.1% vs. 88.4% good outcome, p < 0.001) compared to those without PSIP. Additionally, patients with isolated psychiatric symptoms were younger and had a higher proportion of autoimmune encephalitis (20.8% vs. 9.6%).ConclusionPSIP represents a critical clinical situation in which underlying neurological disorders, particularly viral encephalitis and autoimmune encephalitis are common. Hyponatremia serves as a readily available diagnostic clue and independent predictor of poor prognosis. The absence of neurological signs does not exclude serious underlying pathology. However, these findings are derived from a single tertiary center with a modest sample size and require validation in broader populations. Early comprehensive evaluation, including CSF analysis and neuroimaging, is essential to improve outcomes in this diagnostically challenging population.

BACKGROUND:Mild behavioral impairment (MBI) is an early neurobehavioral marker of dementia, yet MBI domain patterns remain underexplored among populations of Chinese ethnicity. This study aimed to characterize MBI domain phenotypes by examining the prevalence of MBI domains and identifying the leading domain across multi-regional cohorts of dementia-free older adults of Chinese ethnicity. METHODS:Data from three previously unpublished datasets (Hangzhou community cohort, China Longitudinal Aging Study and Singapore memory clinic cohort) and three published studies were integrated to estimate the MBI domain prevalence, measured by the Neuropsychiatric Inventory (NPI) and/or MBI-Checklist (MBI-C), through a random-effects meta-analysis. Within the Hangzhou cohort, cross-instrument consistency was evaluated. Exploratory analyses were performed in the Singapore cohort on associations between MBI domains and incident dementia. RESULTS:Among 1817 participants, impulse dyscontrol was the most prevalent MBI domain, followed by affective dysregulation and decreased motivation, consistently across instruments and cognitive status. In the exploratory longitudinal analyses, impulse dyscontrol was associated with a greater likelihood of incident dementia (HR = 5.05, 95%CI = 2.92 - 8.73). CONCLUSIONS:Impulse dyscontrol was the leading MBI domain among older adults of Chinese ethnicity, with potential clinical relevance for early identification and dementia risk stratification.

Background Environmental pollutant exposure has been associated with chronic kidney disease (CKD), yet specific chemicals related to renal function variation remain incompletely characterized. Methods We analyzed the plasma exposome using liquid chromatography-mass spectrometry (LC-MS) in 734 biopsy-confirmed CKD patients (452 IgA nephropathy [IgAN] and 282 membranous nephropathy [MN]), quantifying 62 chemical exposure species. Associations between exposures and estimated glomerular filtration rate (eGFR) were assessed using regression models and were further evaluated in an independent external cohort (Shanghai Brain Aging Study, SBAS) of 464 older adults. In addition, clinical remission status was assessed after a median follow-up of one year, and its associations with specific chemicals were subsequently evaluated. Proteomic data were integrated to explore protein features related to the exposure-eGFR association in the MN subgroup. Results In the discovery CKD cohort, two PAHs, biphenyl and 2,6-dimethylnaphthalene, were associated with lower eGFR after multivariable adjustment, a finding also observed in the external SBAS cohort. Moreover, in the MN subgroup, these two PAHs were also associated with remission status. Proteomic profiling in MN patients identified six proteins (CDH1, CALR, PTGDS, DAG1, PEBP4, and YIPF3) with statistically significant interaction terms for the biphenyl-eGFR association. Conclusions Combining exposome and proteome analysis, this study identified exploratory associations between PAHs and renal function in CKD patients, and these effects may be related to certain key proteins.

Aims Prediabetes is biologically heterogeneous, but molecular subtypes linked to diabetes progression remain poorly defined. We aimed to identify plasma proteome-based subtypes of impaired fasting glucose (IFG), characterise their molecular features and assess their association with future diabetes risk.Materials and Methods We quantified 2584 plasma proteins using liquid chromatography-mass spectrometry in 538 IFG participants from a prospective discovery cohort (Nutrition and Health of Aging Population in China, NHAPC). Proteomic subtypes were defined by consensus clustering, linked to longitudinal changes in insulin sensitivity and incident type 2 diabetes mellitus (T2DM), which were further validated in an independent Shanghai Brain Aging Study (SBAS) cohort.Results Two reproducible IFG molecular subtypes based on plasma proteomics were identified. The high-risk subtype showed higher incident diabetes and a greater 6-year decline in insulin sensitivity and was characterised by enrichment of glycolysis/gluconeogenesis, insulin signalling and neutrophil degranulation, together with a dyslipidemic lipidomic profile indicating co-dysregulation of glucose and lipid homeostasis. The low-risk subtype demonstrated a higher complement cascade and high-density lipoprotein particle remodelling signature. In the high-risk subtype, key proteins and lipids showed stronger associations with longitudinal declines in insulin sensitivity, including PPBP, PGK1 and ALDOA, as well as PE-P 18:0/20:3 and PE-P 18:1/20:3.Conclusions Proteome-based molecular subtyping stratifies IFG individuals with similar fasting glucose levels but distinct biology and future diabetes risk, supporting earlier and more targeted prevention.

BackgroundIrritability is increasingly recognized for its association with cognitive function, though its impact on cognitive decline and underlying mechanisms remain unclear.ObjectiveTo investigate the associations between irritability and cognition, identify potential neurobiological mechanisms.MethodsThis study included three cohorts: the Alzheimer's Disease Neuroimaging Initiative (ADNI, N = 722), the UK Biobank (UKB, N = 405,112), and the China Longitudinal Aging Study (CLAS, N = 240). Participants were classified into irritability-positive (+) and irritability-negative (-) groups based on assessment of irritability.We used Linear mixed-effects models to assess irritability-related cognitive trajectory, Cox regression to estimate cognitive decline, mediation analysis to test the effect of amyloid-β (Aβ) on the relationship between irritability and cognitive decline, and enrichment analysis to identify the underlying pathological mechanisms of irritability.ResultsIrritability was associated with increased cognitive decline in both ADNI (HR = 1.49, 95% CI: 1.12-1.98) and UKB (HR = 1.09, 95% CI: 1.04-1.15) cohorts, with baseline irritability linked to faster Mini-Mental State Examination decline (2.76 versus 1.88). Mediation analysis showed that cerebrospinal fluid (CSF) Aβ mediated 19-24% of irritability's effect on cognitive decline, while precuneus Aβ pathology mediated 30-41%. Imaging analysis revealed significant thinning of the left precuneus cortex in individuals with irritability. Proteomic analysis indicated underlying pathways involving enhanced energy metabolism and suppressed signal transduction, with modifiable factors (air pollution and physical inactivity) associated with irritability-related pathological proteins.ConclusionsOur findings indicate that irritability is significantly associated with cognitive decline. This association may be driven by mechanisms involving precuneus pathology, increased energy metabolism, and suppressed signal transduction, though these results warrant confirmation in future studies.

Research into an effective dementia treatment is ongoing. Therefore, identifying individuals at risk of declining cognition and dementia is fundamental for initiating modifiable risk factor interventions that can delay dementia onset. Research into modifiable risk factors has almost exclusively been from high-income countries, despite 60% of individuals with dementia living in low- and middle-income countries (LMICs). Addressing this research inequality, the current study examines cross-sectional relationships between risk factors and cognitive performance in LMICs, with the aim of identifying modifiable risk factors particularly suitable for interventions in these regions. Data were obtained from 15 members of the Cohort Studies of Memory in an International Consortium (COSMIC), representing 11 countries (Brazil, China, Colombia, Cuba, India, Indonesia, Malaysia, Nigeria, Republic of Congo, Tanzania, & Uganda) across 6 continents (participants: 53,136; M age  = 70.75, SD age  = 7.87; 57% female). We investigated (after harmonisation) the following risk factors: age, APOE ε4, anxiety, body mass index (BMI), blood pressure, cardiovascular disease, cholesterol, depression, diabetes, education, excessive alcohol, hypertension, hearing loss, physical activity, sex, smoking status, and stroke history. Harmonised global cognition was the outcome. On the AD Workbench, we performed linear regressions for each risk factor within each study at baseline. Cross-sectional results from all studies were pooled in a multivariate meta-analysis, with a random intercept for Country and study. We investigated age, sex and education as risk factors, and included them as covariates when analysing other factors. Older age (β= - .231, p  < .001), being male (β=.151, p  < .001), less education (β=1.569, p  < .001), history of angina (β= - .099, p  < .001), anxiety (β= - .262, p  < .001), depression (β= - .203, p  < .001), stroke (β= - .375, p  < .001), diabetes (β= - .060, p  = .002), excessive alcohol consumption (β= - .190, p  < .001), currently smoking (β= - .097, p  < .001), and less physical activity (β=.160, p  < .001) were significantly associated with worse global cognition. Heterogeneity in the relationship between risk factors and cognition exists across countries, warranting further analyses for each LMIC. Findings suggest that risk factor interventions may help to reduce dementia in LMICs, though country level heterogeneity in the effects suggests that tailoring interventions to regions will be needed. Future research will explore how the factors we identified predict incidence of dementia in LMICs using Machine Learning models.

Background Although poor sleep is widely assumed to impair cognitive function, the impact of sleep disturbances (SD) on language function and the underlying mechanisms of this relationship remains unclear.Objective This study aimed to investigate the association between SD and language function in non-demented elderly individuals, identify potential neuroimaging correlates, and analyze risk factors for SD.Methods We analyzed 784 non-demented elderly subjects from the Alzheimer's Disease Neuroimaging Initiative (ADNI), categorized into SD (n = 256) and normal sleep groups (n = 528) based on self-reported sleep status. Cognitive differences were assessed, and the findings were validated using the China Longitudinal Aging Study (CLAS) and the Chinese Longitudinal Healthy Longevity Survey (CLHLS). Diffusion tensor imaging (DTI) metrics were correlated with language function, and SD risk factors were examined.Results In the ADNI cohort, elderly individuals with SD exhibited worse language function compared to those with normal sleep, and this finding was validated in the CLAS and CLHLS cohorts. Meanwhile, the decline in longitudinal language function among elderly individuals with SD occurred at a faster rate. Differences in DTI metrics between the two groups were primarily observed in the limbic and prefrontal regions. Finally, the risk factors for elderly with SD mainly included years of education, physical and emotional conditions, lifestyles, living environment, and parental survival status.Conclusions SD correlates with language impairment in non-demented elderly, possibly due to limbic/prefrontal tract damage. Risk factors encompass demographic, health, lifestyle, and socio-environmental aspects. Effectively managing these factors and treating SD may improve language function.

About 50-90% people with dementia would develop behavioral disturbances, namely, behavioral and psychological symptoms of dementia (BPSD). Antipsychotic medications are widely used to control severe BPSD symptoms which suffers serious safety risks. It is challenge for individualized precise prediction of antipsychotic drug doses. Neuroimaging, particularly MRI, reveals brain structure associated with aging, cognitive decline, and psychiatric symptoms, making it a potential tool for predicting the drug doses. This study employs transfer learning to predict drug dose and offer neuroanatomical interpretation of BPSD from the perspective of deep learning. We employed a two-step process to train our model (Figure 1). Initially, a large dataset from the Chinese Brain Molecular and Functional Mapping (CBMFM) project (n=646, 334 females and 312males, age 18-82) was used to pretrain the model with a brain age prediction task. Subsequently, the pretrained model was fine-tuned for drug dose prediction from the Alzheimer's Disease and Related Disorders Center in Shanghai Jiao Tong University (ADRDC) dataset (83 BPSD patients, 27 males and 56 females, age 55-80). Finally, we utilized gradient-weighted class activation mapping to generate attention maps and conducted statistical analyses on the attention maps to identify critical brain regions for drug dose prediction. To determine the individual usage of different antipsychotic drugs, the concept of defined daily dose (DDD) was used. The DDD, which individualized control the BPSD, was calculated, ranging from 0 to 1.5 mg/day, serving as the label for fine-tuning. Our pretrained Cas-ResNet exhibited enhanced performance with fewer training epochs, achieving a competitive Pearson correlation of 0.59 between estimated and real DDD (Figure 2). Through feature interpretability analysis, we identified five significant brain regions crucial for BPSD drug dose prediction, mainly located in the temporal lobe, including the parahippocampal area and the striatum (putamen and caudate)(Figure 3). These findings indicate that the antipsychotic dosage to control the BPSD is linked to brain structural alterations, involving both dementia-related and emotion-regulating areas. For the first time, we showed a promising result of using a lightweight deep learning model to predict drug dose prescribed for controlling BPSD. The work promotes the discussion toward appropriate use of antipsychotics in patients with dementia.

INTRODUCTION:This study aimed to comprehensively assess the impact of caregiver, environmental, and individual factors on agitation symptoms in patients with Alzheimer's disease (AD) and identify key modifiable factors. METHODS:From October 2022 to June 2023, 220 participants (110 patients with AD and their caregivers) were recruited from the Shanghai Mental Health Center. Patients with AD completed demographic, lifestyle, medical history, and neuropsychological tests, such as the Mini-Mental State Examination (MMSE) and Geriatric Depression Scale (GDS). The caregivers completed the Neuropsychiatric Inventory, environmental factor questionnaire, and emotional state assessments (Hamilton Depression and Anxiety Scales). Agitation severity was assessed using the Cohen-Mansfield Agitation Inventory (CMAI). Group differences and relationships between potential factors and agitation were also analyzed. RESULTS:Among the 110 patients with AD, 56.36% exhibited agitation. The agitation group had more male patients (p = 0.012) and female caregivers (p = 0.003), lacked courtyard/garden views (p = 0.007), and had lower MMSE (p = 0.005) and GDS (p = 0.012) scores. After adjusting for variables, access to rooms with courtyard/garden views (OR = 0.256, p = 0.042), male caregivers (OR = 0.246, p = 0.005), and higher MMSE scores (OR = 0.194, p = 0.007) were protective factors against agitation. Male caregivers' presence was associated with a lower incidence of agitation. CONCLUSIONS:Improving living environments, promoting male caregivers, enhancing caregiver support, and early cognitive intervention may reduce agitation in AD patients.

Social support is considered a protective factor against depression, but there are inconsistent findings regarding social support and depression in older adults. We aimed to clarify the association between emotional and instrumental social support and depressive symptoms in older adults cross-sectionally and longitudinally (mean follow-up = 1.96 years). We meta-analyzed raw individual participant level data from adults in mid- and late life (N = 23 973) who completed questionnaires about physical health, mental health, and social support and completed neuropsychological assessments. These were COSMIC (Cohort Studies of Memory in an International Consortium) cohort studies carried out in Australia, Brazil, China, Germany, Greece, India, Indonesia, Singapore, South Korea, Sweden, and the United States in mostly urban settings. After controlling for depression risk factors, emotional support (B = -0.40 [95% CI, -0.60 to -0.21]), but not instrumental support (B = 0.17 [95% CI, -0.26 to 0.59]), was associated with lower depressive symptoms cross-sectionally and at follow-up [emotional support (B = -0.37 [95% CI, -0.54 to -0.20]); instrumental support (B = 0.09 [95% CI, -0.30 to 0.49])]. Emotional support was associated with lower depressive scores cross-sectionally and longitudinally, while instrumental support was not associated with depressive symptoms. Our findings can help inform the nature of interventions to prevent and reduce risk of depression among older adults. This article is part of a Special Collection on Cross-National Gerontology.

BACKGROUND:Alzheimer's disease (AD) frequently co-occurs with depressive symptoms, exacerbating both cognitive decline and clinical complexity, yet the neural substrates linking this co-occurrence remain poorly understood. We aimed to investigate the role of basal forebrain-limbic system circuit dysregulation in the interaction between cognitive impairment and depressive symptoms, identifying potential biomarkers for early diagnosis and intervention. METHODS:This cross-sectional study included participants stratified into normal controls (NC), cognitive impairment without depression (CI-nD), and cognitive impairment with depression (CI-D). Multimodal MRI (structural, diffusion, functional, perfusion, iron-sensitive imaging) and plasma biomarkers were analyzed. Machine learning models classified subgroups using neuroimaging features. RESULTS:CI-D exhibited distinct basal forebrain-limbic circuit alterations versus CI-nD and NC: (1) Elevated free-water fraction (FW) in basal forebrain subregions (Ch123/Ch4, p < 0.04), indicating early neuroinflammation; (2) Increased iron deposition in the anterior cingulate cortex and entorhinal cortex (p < 0.05); (3) Hyperperfusion and functional hyperactivity in Ch123 and amygdala; (4) Plasma neurofilamentlightchain exhibited correlated with hippocampal inflammation in CI-nD (p = 0.03) but linked to basal forebrain dysfunction in CI-D (p < 0.05). Multimodal support vector machine achieved 85 % accuracy (AUC=0.96) in distinguishing CI-D from CI-nD, with Ch123 and Ch4 as key discriminators. Pathway analysis in the CI-D group further revealed that FW-related neuroinflammation in the basal forebrain (Ch123/Ch4) indirectly contributed to cognitive impairment via structural atrophy. CONCLUSION:We identified a neuroinflammatory-cholinergic pathway in the basal forebrain as an early mechanism driving depression-associated cognitive decline. Multimodal imaging revealed distinct spatiotemporal patterns of circuit dysregulation, suggesting neuroinflammation and iron deposition precede structural degeneration. These findings position the basal forebrain-limbic system circuit as a therapeutic target and provide actionable biomarkers for early intervention in AD with depressive symptoms.

Mild cognitive impairment (MCI) is a critical stage in dementia prevention. This study aims to explore key modifiable risk factors for MCI among community-dwelling elderly in Shanghai, China, by incorporating potential factors unique to China into the established list of modifiable risk factors at an earlier stage. 1000 participants with a baseline diagnosis of normal cognition (NC) were selected and involved in this study through the Shanghai Brain Health Cohort Study. All the participants completed a self-administered questionnaire collecting demographic, lifestyle, and clinical information, as well as a comprehensive neuropsychological assessment at both baseline and each two-year follow-up visit. For continuous variables among the 7 categories of modifiable risk factors that were statistically significant in the univariate COX regression analysis, the Maximally Selected Log-Rank Statistic was used to determine the optimal cutoff point based on the KM curve, converting them into binary variables for Population Attributable Fraction (PAF) calculation. Participants had a mean age of 68.31±7.07 years, with 65.5% being female. During the follow-up period, 267 individuals (26.7%) developed MCI. Seven modifiable risk factors were identified: being a manual laborer (HR = 1.02, p  = 0.042), abnormal marital status (divorced or widowed) (HR = 1.49, p  = 0.016), comorbidity including diabetes (HR = 1.73, p  < 0.001) and cerebral infarction (HR = 1.51, p  = 0.007), depression (HR = 1.03, p  = 0.03), chronic sleep disorder (HR = 1.03, p  = 0.03), and abnormal sleep duration in both youth (HR = 0.89, p  = 0.026) and middle age (HR = 0.87, p  = 0.008). Additionally, eating soy products was found to be protective (HR = 0.75, p  = 0.031). The overall weighted population attributable fraction (PAF) for these risk factors was 24.17%, which increased to 24.39% after adjusting for age and years of education. These findings suggest that comorbidity with diabetes and cerebral infarction, manual labor status, abnormal sleep duration, and not consuming soy products are major modifiable risk factors for MCI in the study population, offering novel insights and potential new research directions for the early prevention of dementia.

Previous studies have found some cognitive benefits from ginger consumption, but there are little data on this among older Chinese. To explore the relationship between ginger consumption and dementia and explore the possible mechanism of ginger consumption on cognitive decline. A total of 410 elderly patients with dementia and 2426 non-dementia individuals were analyzed using data from the Shanghai Brain Health Foundation. Each participant's cognitive diagnosis was made by an attending psychiatrist, and their overall cognitive function was assessed by Montreal Cognitive Assessment (MoCA). The Food Frequency Questionnaire (FFQ) was used to investigate their consumption of ginger. To explore the possible mechanisms of ginger prevention of dementia, 408 non-dementia patients (331 ginger consumers and 77 non-ginger consumers) completed head MRI and plasma Alzheimer's disease (AD) biomarkers such as amyloid-beta peptides (A beta) 42, A beta 40, total tau (t-tau), phosphorylated tau-181 (p-tau-181), and neurofilament light chain (NfL). The incidence of dementia was found to be reduced by ginger consumption through multiple logistic regression analysis. Compared to non-ginger consumers, ginger consumers had higher MoCA scores and lower plasma NfL and A beta 40 levels. Regression analysis and mediated models then showed that ginger consumption reduced plasma NfL concentrations, affecting overall MoCA scores. Ginger consumption may be a protective factor against dementia in elderly Chinese and may prevent cognitive decline by affecting plasma NfL concentration.

While numerous studies strive to exploit the complementary potential of MRI and PET using learning-based methods, the effective fusion of the two modalities remains a tricky problem due to their inherently distinctive properties. In addition, current studies often face the problem of small sample sizes and missing PET data due to factors such as patient withdrawal or low image quality. To this end, we propose a hybrid multi-modality multi-task learning (HM2L) framework with cross-domain knowledge transfer for forecasting trajectories of SCD progression. Our HM2L comprises (1) missing PET imputation, (2) multi-modality feature extraction for MRI and PET feature learning with a novel softmax-triplet constraint, (3) attention-based multi-modality fusion of MRI and PET features, and (4) multi-task prediction of category labels and clinical scores such as Mini-Mental State Examination (MMSE) and Geriatric Depression Scale (GDS). To handle problems with small sample sizes, a transfer learning strategy is developed to enable knowledge transfer from a relatively large scale dataset with MRI and PET from 795 subjects to two small-scale SCD cohorts with a total of 136 subjects. Experimental results indicate HM2L surpasses several state-of-the-art methods in jointly predicting category labels and clinical scores of subjective cognitive decline. Results show that the MMSE scores of SCD subjects who develop mild cognitive impairment during the 2-year/7-year follow-up are significantly lower than those of subjects who remain stable, while there exists a complex relationship between SCD progression with GDS.

Presence of cardiometabolic multimorbidity (CMM) has been linked to depressive symptoms in adults. The present study aimed to investigate the distinctive mapping between CMM and differential neuropsychiatric subsyndromes among multi-regional and ethnical older adults. The present study included discovery and validation datasets. The discovery dataset consisted of two longitudinal and two cross-sectional studies from Asian older adults from both clinical and community settings. The longitudinal validation dataset was from the UK Biobank (UKB). CMM was defined as the coexistence of ≥2 conditions including hypertension, hyperlipidaemia, diabetes mellitus, stroke, and other cardiovascular diseases (CVD). Subsequently, two CMM patterns, including metabolic (≥2 of hypertension, hyperlipidaemia and diabetes mellitus) and cardio-cerebrovascular multimorbidity (≥2 of hypertension, stroke and CVD) were identified respectively. The neuropsychiatric inventory (NPI) was applied to define four neuropsychiatric subsyndromes in the discovery dataset: psychosis, hyperactivity, affective and apathy. In the validation dataset, the four subsyndromes were defined using previously established methods. Short-term incidence of subsyndromes was defined as the development of neuropsychiatric symptoms (NPS) in recent 1-2 years. Two-step individual participant data was used to establish the cross-sectional association between CMM and NPS. Cox regression models explored baseline CMM patterns and NPS incidence in both discovery and validation datasets. The discovery dataset included 2950 Asian participants for cross-sectional analysis and 967 participants (675 NPS-free at baseline) for longitudinal analysis. The validation dataset included 885 participants (765 NPS-free at baseline) for longitudinal analysis. People with CMM had a higher presence of NPS and a higher risk of 1-year incidence of affective syndrome (hazard ratio [HR] = 3.95, 95%CI = 1.44-10.82). Regarding CMM patterns, metabolic multimorbidity was associated with affective syndrome at baseline (odds ratio [OR] = 1.35, 95%CI = 1.01-1.81), as well as its short-term incidence (HR = 3.88, 95%CI = 1.53-9.85). In the validation dataset, CMM, especially metabolic multimorbidity, was also associated with an increased risk of short-term incidence of affective subsyndrome (HR = 8.04, 95%CI = 2.65-24.40, and HR = 6.44, 95%CI = 2.11-19.62, respectively). Cardio-cerebrovascular multimorbidity was not associated with incidence of any NPS in both datasets. CMM, especially metabolic multimorbidity was associated with affective disturbances in multi-regional-ethnical older adults. Our findings emphasized the importance of targeted management of metabolic multimorbidity.