Journal of Neurogenetics

Spousal caregivers of individuals with Alzheimers disease and related dementias frequently experience elevated perceived stress, caregiver burden, and loneliness, which are associated with adverse health outcomes. Early identification is therefore critical for timely intervention. Existing approaches commonly rely on wearable sensor data and standardized psychological questionnaires, while recent multimodal methods aim to improve prediction by integrating behavioral and linguistic information. In this study, we explored three modality configurations, wearable-derived features, interview-based text, and their combination, to classify caregiver psychological risk using the Perceived Stress Scale (PSS), Zarit Burden Interview, and UCLA Loneliness Scale. We compared traditional machine learning models and large language models (LLMs) (Gemini 2.0, Llama 4, and GPT-4o) under psychometrician-centered and caregiver-centered prompting strategies. Traditional machine learning models performed better under multimodal settings, while LLMs achieved stronger performance with Interview-Only input. We further demonstrate that PSS was the most predictable construct and prompting strategies substantially influenced LLM performance.

Sleep gets worse with age and is correlated with risk for disease and mortality. The possibility that poor sleep causes aging to accelerate has prompted interest in improving sleep to slow aging and prevent disease. However, the existing evidence on the link between poor sleep and accelerated aging is unclear. Here, we tested for correlation and causation between poor sleep and accelerated aging using five independent datasets of adults (total N > 64,000). We found strong evidence for a correlation between poor sleep and fast aging that is consistent across young, middle, and late adulthood and across aging biomarkers derived from different tissues and modalities. We found that this correlation is robust to the influence of chronic disease burden, but not to the influence of shared genetic and early environmental factors among twins. Finally, we found mixed evidence for a causal influence of poor sleep on accelerated aging using Mendelian randomization. Our findings indicate that the correlation between poor sleep and accelerated aging is highly robust; however, the claim that poor sleep causes aging to accelerate is not consistently supported.

Behavior arises from the complex interplay between an organisms nervous system, its genetic makeup, and the environment. High-resolution, high-throughput behavioral quantification is essential for dissecting biological function and the effects of genetic perturbation, but automated analysis remains challenging. Here, we present Autobehaver, an automated behavioral analysis pipeline based on a low-cost, high-throughput recording platform that captures videos of individual Drosophila. From each video, we extracted keypoints and used a custom Transformer to assign frame-wise behavior and orientation labels. We then converted these predictions into high-dimensional per-animal feature vectors and trained XGBoost ensembles to classify animals and identify the features that separated groups. By applying SHAP analysis to the classifier ensemble, we identified the behavioral features most informative for distinguishing groups of flies. We demonstrated the approach in several ways. First, we recovered known behavioral changes associated with heat-activated dTrpA1 activity in specific neural circuits. Second, we detected age-associated behavioral changes consistent with gradual impairment of locomotor and climbing ability. Finally, we used Autobehavers classifier ensemble to place animals with intermediate phenotypes along a behavioral axis and used feature-importance analysis to reveal the behavioral features underlying those intermediate states. Together, Autobehaver provides an interpretable framework for quantitative behavioral phenotyping and comparative analysis of complex genotypes.

Introduction: Bipolar disorder (BD) has been associated with increased medical burden and accelerated biological aging. Long non-coding RNAs (lncRNAs) regulate molecular pathways related to cellular senescence, inflammation, and telomere maintenance, which are implicated in both BD and aging. This study examined whether aging-related lncRNA expression reflects familial vulnerability or illness-specific effects, and whether childhood trauma and lifestyle factors modulate these signatures within a gene-environment framework. Methods: In this cross-sectional study, expression levels of aging-related lncRNAs, including Antisense Non-coding RNA in the INK4 Locus (ANRIL), HOX Transcript Antisense Intergenic RNA (HOTAIR), Nuclear Enriched Abundant Transcript 1 (NEAT1), Taurine Upregulated Gene 1 (TUG1), Metastasis-Associated Lung Adenocarcinoma Transcript 1 (MALAT1), Growth Arrest-Specific 5 (GAS5), and Telomerase RNA Component (TERC), were measured in peripheral blood mononuclear cells (PBMCs) from individuals with bipolar disorder (BD) (n=68), siblings without BD diagnosis (SIB) (n=54), and healthy controls (HC) (n=70) using quantitative reverse transcription polymerase chain reaction (RT-qPCR). Childhood trauma and lifestyle were assessed using the Childhood Trauma Questionnaire (CTQ) and the Healthy Lifestyle Profile II (HPLP-II). Principal component analysis generated a composite aging-related lncRNA factor. Results: At the individual transcript level, NEAT1 and TERC were elevated, whereas GAS5 was reduced, in both BD and SIB relative to HC. The aging-related lncRNA composite score was higher in BD and SIB than in HC (F = 7.315, p = 0.001). Familial liability to BD (presence vs. absence of familial liability) showed a significant main effect on the composite score (F(1,182)=8.18, p=0.005) and interacted with childhood trauma (F(1,182)=10.14, p=0.002). In multivariable models, total CTQ and physical neglect were independently associated with lower composite scores, while familial liability remained a positive predictor (all p<0.001). Conclusions: Aging-related lncRNA alterations mark familial vulnerability to BD and are shaped by childhood trauma within a gene-environment interaction framework.

Gamma oscillations (30-100 Hz) are critical for cognitive processing, and their disruption is associated with Alzheimers disease (AD) and related dementias. Gamma ENtrainment Using Sensory Stimulation (GENUS) therapy applies 40 Hz light and/or sound to restore gamma oscillations, but clinical trials report highly variable responses. Using Kuramoto oscillator models calibrated to human neurophysiology, we demonstrate that frequency mismatch between stimulation frequency and individual intrinsic gamma frequency (IGF) is a critical determinant of entrainment efficacy. A 2 Hz frequency mismatch reduced phase-locking value (PLV) by 43-82%, depending on neural resonance bandwidth, with PLV at mismatch indistinguishable from the finite-sample noise floor -- indicating complete absence of stimulus-synchronized oscillation. In a simulated population with IGF drawn from a clinically realistic distribution (mean 35.10 {+/-} 3.49 Hz, N = 200), fixed 40 Hz stimulation achieved mean PLV of 0.119 {+/-} 0.081, compared to 0.504 {+/-} 0.009 for individualized frequency stimulation -- a 4.2-fold advantage (t(199) = 67.26, p < 0.001, Cohens d = 4.76). In the clinically relevant subgroup with IGF < 36 Hz (62% of the population), the fold advantage increased to 5.3x. Stochastic noise sensitivity analysis confirmed robustness of the fold advantage (3.8-4.0x across{sigma}{square}{square}{square}{square}{square}{square} = 0-2.0 rad/s; p < 10{square}3{superscript 1} at all levels). These findings provide quantitative computational support for personalized GENUS protocols incorporating individual gamma frequency measurement and carry direct implications for the design of next-generation clinical trials.

Animals sleep more when they are sick. In C. elegans, stress-induced sleep (SIS) follows cellular injury such as exposure to ultraviolet (UV) light. The genetic regulators of SIS remain incompletely defined. Using a worm-picking robot, multi-well WorMotel imaging, and association analysis we performed a semi-automated screen of 941 whole-genome-sequenced Million Mutation Project (MMP) strains. We quantified behavioral activity and quiescence before and after ultraviolet (UV) radiation. We applied the Sequence Kernel Association Test (SKAT) to this behavioral data to prioritize 6,663 genes and observed significant enrichment of known SIS genetic regulators. Based on these results, we conducted a candidate validation screen for additional genes regulating SIS. We identified three genes (strd-1, egl-8, cla-1), mutations in which reproducibly influence SIS. Further exploration of these genes holds potential for enhancing our understanding of the molecular basis of SIS. These findings establish a pipeline for automated behavioral phenotyping coupled with gene-based association to accelerate studies of C. elegans neurogenetics.

ImportanceAdverse neighborhood conditions can lead to poorer health outcomes, potentially through accelerated biological aging. However, whether these relationships are explained by individual- or neighborhood-level factors remains unclear. ObjectiveTo examine the association between neighborhood deprivation, measured by the Area Deprivation Index (ADI), and epigenetic age acceleration and assess whether individual- and neighborhood-level characteristics mediate or modify these associations. DesignCross-sectional study using data from a Yale Stress Center study between 2008 and 2012. Data analysis was conducted from July 2025 to January 2026. SettingCommunity-based sample from the greater New Haven, CT area. ParticipantsA total of 370 healthy adults aged 18 to 50 years without major psychiatric, medical, or cognitive disorders who provided blood samples for DNA methylation analysis. Main Outcomes and MeasuresEpigenetic age acceleration measured from DNA methylation using four second-generation epigenetic clocks, with associations assessed among aging, neighborhood deprivation, and individual- and neighborhood-level factors. ResultsData were analyzed from 370 participants (212 women [57.3%], 158 men [42.7%]; mean [SEM] age, 29.3 [0.46] years). Greater neighborhood deprivation was associated with greater lifetime adversity ({beta}=0.112, p<.001) and lower educational attainment ({beta}=-0.019, p=.012), and accelerated epigenetic aging as measured by GrimAge ({beta}=0.037, p<.001), PCGrimAge ({beta}=0.019, p<.001), and PCPhenoAge ({beta}=0.041, p<.001), but not PhenoAge (p=.23). In multivariable models accounting for individual factors, neighborhood deprivation remained associated with these three clocks. Lifetime adversity partially mediated the association between ADI and accelerated GrimAge (20.3% of total effect) and PCGrimAge (23.3%). Race moderated the direct association between ADI and epigenetic aging, with stronger associations between neighborhood deprivation and accelerated GrimAge ({beta}=0.061, p=.004) and PCPhenoAge ({beta}=0.057, p=.02) observed among Black participants compared to White. ConclusionsGreater neighborhood deprivation was associated with accelerated epigenetic aging across multiple second-generation clocks, with lifetime adversity partially mediating these associations. Stronger effects were observed among Black participants. These findings suggest that neighborhood environments and cumulative stress may contribute to biological aging and racial disparities in aging trajectories. Key PointsO_ST_ABSQuestionC_ST_ABSIs neighborhood deprivation associated with epigenetic age acceleration, and if so, how do neighborhood- and individual-level factors impact this relationship? FindingsIn this cross-sectional study of 370 adults, greater neighborhood deprivation was associated with accelerated epigenetic aging across multiple second-generation clocks. Lifetime adversity partially mediated these associations, and the relationship between neighborhood deprivation and accelerated aging was stronger among Black participants than White participants. MeaningThese findings suggest that neighborhood conditions and lifetime stress contribute to accelerated biological aging and suggest that epigenetic aging may represent one biological pathway through which neighborhood-level racial inequalities contribute to health disparities.

Sleep-wake regulation arises from the interaction between homeostatic sleep pressure and circadian timing, yet current assessments evaluate these processes independently and fail to capture their dynamic modulation by environmental pressures. This limitation is particularly relevant in adolescents with attention-deficit/hyperactivity disorder (ADHD), who are at increased risk of circadian delay and sleep disruption. Here, we combined month-long wearable-based physiological monitoring with ecological behavioral assessments in adolescents with ADHD to characterize circadian and homeostatic processes dynamically in real-world settings. Using continuous skin temperature recordings, we derived individualized and day-specific estimates of circadian phase through hierarchical modelling, and integrated these measures with actigraphy-based sleep estimates and daily assessments of neurocognitive functioning and functional impairment. Temperature-derived circadian phase correlated with questionnaire-based chronotype but more accurately predicted sleep patterns. Delayed circadian phase was associated with later sleep onset and greater weekday-weekend variability. Importantly, circadian phase exhibited significant day-to-day fluctuations, particularly in individuals with delayed phase, reflecting interactions with environmental constraints. Sleep latency was jointly determined by homeostatic sleep pressure and day-specific circadian phase, with combined models outperforming either process alone. Crucially, both sleep deprivation and day-specific circadian misalignment independently predicted fluctuations in ADHD symptom severity, perceived stress, and neurocognitive impulsivity. In contrast, mean circadian phase alone did not explain behavioral variability. These findings demonstrate that circadian regulation is a dynamic, environmentally sensitive process rather than a fixed trait. Wearable-based estimation of circadian phase provides a scalable approach to capture these dynamics and may enable personalized interventions targeting sleep and circadian dysregulation.

Dominant missense mutations in ATP1A3, encoding a Na+, K+ ATPase -3 subunit, can cause Alternating Hemiplegia of Childhood (AHC), but how these mutations lead to AHC remains unclear. Here, we establish the first C. elegans AHC models by introducing AHC-causing ATP1A3 patient mutations (D801N, E815K, L839P, and G947R) into the orthologous gene, eat-6, using CRISPR/Cas9. Homozygous C. elegans AHC model animals have recessive developmental defects. Heterozygous AHC model animals have dominant defects in neuromuscular junction (NMJ) function that are inconsistent with haploinsufficiency and dominant sleep or arousal defects. Previous work in a Drosophila G755S AHC model found that loss of a K-dependent, Na/Ca{superscript 2} exchanger exacerbated neuronal defects. We introduced a loss-of-function allele of the orthologous C. elegans gene, ncx-4, into C. elegans AHC models; loss of ncx-4 function did not consistently alter C. elegans AHC model defects across alleles. Our results establish novel C. elegans models of AHC with robust phenotypes, demonstrate that AHC mutations disrupt NMJ function, and provide proof-of-concept for discovering cross-species modifiers of AHC-related phenotypes. Summary StatementWe report the first C. elegans models of Alternating Hemiplegia of Childhood. D801N, E815K, L839P, and G947R AHC model animals have recessive development defects and dominant neuromuscular defects.

Animals balance staying in a favorable environment with exploring new ones. In C. elegans chemotaxis, the process by which worms migrate toward an attractant has been extensively studied. However, what happens after they reach it remains largely unexplored, partly because conventional assays immobilize worms at the point of arrival. Here, we quantitatively analyzed chemotactic behavior upon reaching an attractive odor source using an immobilization-free chemotaxis assay. We observed that 62% animals left the isoamyl alcohol region after initially approaching it, a behavior we termed "leaving behavior." Quantitative analysis revealed that leaving behavior represents a distinct locomotor state compared with free-moving, high-concentration odor avoidance, and approach behavior. To test whether leaving behavior is related to olfactory adaptation, we analyzed mutants in adaptation-related genes. The proportion of leaving behavior was significantly increased in egl-4 loss-of-function mutants compared with wild-type animals, whereas arr-1 mutants showed no significant difference. These results suggest that egl-4 negatively regulates leaving behavior, suggesting a role for this kinase in stabilizing post-arrival behavioral states beyond its known function in olfactory adaptation. Our findings indicate that chemotaxis involves dynamic behavioral transitions even after reaching an attractant, consistent with an exploration-exploitation trade-off framework.

Octopamine is involved in a variety of different physiological and behavioral mecha-nisms in Drosophila melanogaster. Throughout the life cycle of the fruit fly, from the larva to the adult, octopaminergic neurons in both the central and the peripheral nerv-ous system target a multitude of neurons and even non-neuronal tissues, making it challenging to analyze individual mechanisms of octopamine function. One approach to deconstructing this complex system is to examine the postsynaptic components of signal transmission. In Drosophila, octopamine interacts with six distinct G-protein-coupled receptors. For some of these receptors, expression maps and functional im-plications have been described. In contrast, other receptors have been neglected, partly due to the lack of suitable genetic tools. Here, for the first time, we compiled a complete set of mutant lines of all known octopamine receptors, all generated using the same genetic tool, the recently established Trojan Exon system. It integrates the Gal4/UAS binary expression strategy while simultaneously impairing receptor func-tion. This enabled us to generate a comprehensive anatomical map of receptor ex-pression in the larva and, at the same time, analyze the function of individual octopa-mine receptors during larval development, chemosensory perception and locomotion. All octopamine receptors (Oamb, Oct2R, Oct{beta}1R, Oct{beta}2R, Oct{beta}3R, and Oct-TyrR) showed extensive signal in the central nervous system. The same was found for the peripheral nervous system, with the exception of Oct{beta}2R, which showed pronounced expression in the somatic muscles. We also observed a previously undescribed role of Oct{beta}1R, Oct{beta}3R, and Oct-TyrR in larval hatching and in the survival of larvae and pupae. Molecular evaluation of the Trojan Exon octopamine lines supports our analy-sis. In addition, we combined the experimental results with gene expression data from the different development stages of Drosophila melanogaster and from different tis-sues and cell populations throughout the body. Overall, we compiled, analyzed and validated a complete set of octopamine lines which, together with gene expression analysis, provides a basis for further functional studies on the larval octopaminergic system.

Background and ObjectivesGamma Entrainment responses in Alzheimers disease are heterogeneous, but factors underlying this variability remain unclear. Cognitive reserve may influence how large-scale brain networks reconfigure during stimulation, particularly in prodromal Alzheimers disease. We examined whether cognitive reserve was associated with stimulation-related functional network changes across the Alzheimers disease spectrum and hypothesized a condition-by-diagnosis-by-cognitive reserve interaction. MethodsIn this cross-sectional EEG study, adults aged 55 years or older were recruited from a memory clinic and community. Participants underwent amyloid PET and standardized neuropsychiatric and neuropsychological assessment and were classified as amyloid-negative cognitively normal, amyloid-positive mild cognitive impairment, or amyloid-positive Alzheimers disease dementia. Amyloid-negative mild cognitive impairment and amyloid-positive cognitively normal participants were excluded from the primary analyses. Cognitive reserve was indexed using an education- and occupation-based composite and dichotomized at the median. Functional networks during rest and 32-Hz visual stimulation were derived from spectral Granger causality EEG and summarized using mean strength, weighted clustering coefficient, average shortest path length, and outreach. Mixed-design repeated-measure ANCOVA adjusted for age, sex, and APOE genotype was used for the primary analysis. ResultsSignificant condition-by-diagnosis-by-cognitive reserve interactions were observed for mean strength, weighted clustering coefficient, and outreach, but not for average shortest path length. These effects were driven primarily by prodromal Alzheimers disease, in which the low cognitive reserve subgroup showed larger rest-to-stimulation increases than the high cognitive reserve subgroup for mean strength, weighted clustering coefficient, and outreach. In secondary analyses, the low cognitive reserve prodromal subgroup also showed higher 32-Hz entrainment, and entrainment strength correlated positively with stimulation-related network change across the full sample. DiscussionCognitive reserve may be associated with heterogeneity in stimulation-related network reconfiguration, with the clearest signal observed in prodromal Alzheimers disease. Because the low-reserve prodromal subgroup was small and findings were sensitive to alternative modeling choices, these results should be considered preliminary. Reserve-related factors may warrant explicit consideration in future studies of sensory gamma stimulation in Alzheimers disease.

Background: Low-dose levetiracetam is under investigation as a potential treatment for slowing Alzheimer's Disease progression. This study tests whether levetiracetam enhances executive function in mid-age adults, and whether drug effects differ by Apolipoprotein e4 (APOE4+) genetic risk status. Methods: Fifty-eight adults (aged 45-65 years; 27 APOE33; 31 APOE4+) participated in a double-blind, placebo-controlled study of low-dose levetiracetam (125mg bidaily for two-weeks). At the end of each treatment phase, participants completed a switch-inhibition task. Results: Mid-age APOE4+ carriers were significantly slower and showed a greater cost of increasing executive demand than APOE33 individuals. Response times were quicker under levetiracetam, with increased benefits reported in APOE33 individuals, at younger ages, and in individuals with reduced levels of plasma-based biomarkers. Levetiracetam selectively benefitted accuracy in APOE33 individuals. Conclusion: Low-dose levetiracetam enhances executive function in midlife, particularly in individuals at lower risk of Alzheimer's Disease based on age, APOE4 genotype, and proxies of neuropathology.

Relatively little is known regarding suicide among older adults in nursing homes. The aim of this study was to compare the incidence of suicide among older nursing home residents (NHR) with community-dwelling older people (CDP) using newly available, national, individual-level register data, and to assess differences with regard to socio-demographic characteristics. We obtained data on all older adults aged 65+ who were living in Austria at the end of October 2018 (n=1,665,450), including 155,020 NHR. Death by suicide was followed until the end of 2023. A total of 114 and 2,136 suicides were observed among NHR and CDP; corresponding to cumulative incidences of 14 and 27 per 100,000, respectively. Among NHR, suicide incidence was higher among males (28.0, 95% CI=22.1, 35.5), those aged 65-74 years (20.2, 95% CI=13.3, 30.6), with tertiary education (23.3, 95% CI=10.6, 50.6), divorced (25.0, 95% CI=16.2, 38.5), and residing in urban nursing homes (22.0, 95% CI=17.0, 28.4). Compared to CDP, more suicides in NHR occurred by poisoning and but few by firearms. In conclusion, we found that suicide incidence was lower among older NHR compared to CDP. More research on and preventive efforts against suicide among older NHR are needed.

Background Greater cognitive intra-individual variability (IIV) reflects increased heterogeneous performance across cognitive domains and has been linked to a higher risk of Alzheimer's disease (AD). However, it remains unclear whether cognitive IIV is linked to heterogeneous dispersion of regional AD pathology. Hence, we aimed to examine the association between cognitive IIV and AD neuroimaging biomarker IIV. Methods This study included participants with normal cognition (CN) and mild cognitive impairment (MCI) from the Alzheimer's Disease Neuroimaging Initiative. Cognitive IIV was computed as the within-person standard deviation of five domain-specific neuropsychological test z-scores. Four neuroimaging biomarker IIV metrics were similarly derived using regional amyloid-{beta} (n = 1,021), tau (n = 719), cortical thickness (n = 2,148), and combined amyloid-tau-neurodegeneration (ATN, n = 258). Associations between cognitive IIV and each biomarker IIV were evaluated using linear regression models, adjusted for relevant covariates. Results Higher cognitive IIV was associated with greater biomarker IIV across amyloid-{beta} ({beta} = 0.039, SE = 0.014, p = .006), tau ({beta} = 0.196, SE = 0.033, p < .001), cortical thinning ({beta} = 0.036, SE = 0.008, p < .001), and ATN ({beta} = 0.176, SE = 0.043, p < .001). Interaction analyses revealed that the associations of cognitive IIV with tau IIV, cortical thickness IIV, and ATN IIV were stronger in MCI than CN individuals. Significant interactions between cognitive IIV and biomarker positivity status showed that the effect with amyloid-{beta} IIV was attenuated in A- ({beta} = 0.004, SE = 0.014, p = .78) but that the effect with tau IIV remained robust even in T- individuals ({beta} = 0.088, SE = 0.022, p < .001). Conclusion Elevated cognitive IIV is associated with greater heterogeneity in cortical dispersion of AD-related pathology, particularly in prodromal AD and in the presence of abnormal pathology. As a novel measure that captures variation in topographical scattering of AD pathological burden across the cortex, AD biomarker IIV may offer research and clinical utility beyond evaluating absolute biomarker load or thresholds.

Background: Depressive and anxiety symptoms are prevalent among older adults and associated with increased dementia risk. In healthy older adults, higher step counts are associated with fewer depressive and anxiety symptoms; whether this holds in individuals with cognitive concerns (subjective cognitive decline [SCD] or mild cognitive impairment [MCI]) is unknown. In a randomised controlled trial, the 12-month APPLE-Tree group psychosocial lifestyle intervention produced small cognitive improvements but no change in step count. Objective: To test whether step count was associated with depressive and anxiety symptoms (cross-sectionally and over 24 months), and whether APPLE-Tree increased step count in participants with clinical anxiety or depression. Methods: We examined cross-sectional and longitudinal (12- and 24-month) associations between step count (two-week average from wrist-worn wearables) and depressive and anxiety symptoms (Hospital Anxiety and Depression Scale) using adjusted linear regressions, with a mediation analysis of self-perceived mobility. We also tested whether the intervention increased step count in those with baseline clinical anxiety or depression. Findings: We included 629 of 746 trial participants at baseline, of whom 376 contributed 12-month and 215 24-month data. At baseline, higher step counts were associated with fewer depressive symptoms ({beta} = -0.11, 95% CI -0.17 to -0.05, p < 0.001) but, counter to our hypothesis, more anxiety symptoms ({beta} = 0.12, 95% CI 0.06 to 0.19, p = 0.003). Over two years, change in step count was not associated with change in depressive or anxiety symptoms (all p [&ge;] 0.12). Self-perceived mobility problems mediated the association between step count and depressive but not anxiety symptoms. The intervention did not change step count in those with clinical anxiety or depression. Conclusions: This provides the first evidence in older adults with cognitive concerns that higher step counts are associated with fewer depressive but more anxiety symptoms. This may reflect heterogeneity of a population that includes those with prodromal dementia and cognitive health anxiety. Step count did not predict symptoms over time. Clinical implication: Step count may help distinguish anxiety and depressive symptoms in people presenting with cognitive concerns, or underlying reasons for cognitive concerns among those with functional cognitive disorders.

Autism spectrum disorder (ASD) has been increasingly linked to gut microbial dysbiosis, yet the developmental trajectory of microbiota and microbial metabolites across childhood remains poorly understood. Here, we performed an integrated multi-omics analysis of gut bacterial taxa and faecal metabolites, including volatile organic compounds (VOCs), short-chain fatty acids (SCFAs), and amino acids (AAs), in 101 individuals with ASD and 105 age-matched neurotypical controls. The study was conducted in a well-powered, age-stratified pediatric cohort including toddlers, children, and adolescents. Our findings reveal early-life microbial dysbiosis in ASD, characterized by elevated alpha-diversity, altered community composition, and age-specific enrichment of taxa including Faecalibacterium, Bacteroides, Ruminococcus, Alistipes, and Roseburia. Network analyses demonstrated that toddlers exhibit smaller, sparser microbial interaction networks, while children and adolescents show increasingly complex and interconnected networks, compared with respective age-matched controls, suggesting a critical developmental window for microbiome-host interactions. Metabolomic profiling identified consistent ASD-associated alterations, including elevated aromatic and indole-derived VOCs (i.e., N-ethyl-benzenamine, 6-methyl-5-hepten-2-one, methyl isobutyl ketone), disrupted SCFA patterns, and reduced D-aspartate and D-alanine, indicating functional reprogramming of microbial fermentation and D-amino acid metabolism. Correlations between specific taxa and metabolites suggest that microbial community structure drives these metabolic outputs, with potential impacts on gut-brain signalling, immune modulation, and neurobehavioral phenotypes. These results support a model in which ASD involves early-life gut microbiota alterations that persist but evolve across development, highlighting potential microbial and metabolic biomarkers. Importantly, they underscore the translational potential of microbiome-targeted interventions during early childhood to modulate neurodevelopmental outcomes.

Motivation: Linkage disequilibrium score (LDSC) regression is a popular method to estimate heritability for complex traits using summary statistics and linkage disequilibrium (LD) reference panels, offering a practical alternative to methods requiring individual-level data. Despite its widespread use, LDSC regression can produce biased heritability estimates. The properties of LDSC regression were investigated using summary statistics from several large-scale Alzheimer's disease (AD) studies and a variety of LD reference panels. These heritability estimates were compared with those obtained from individual-level data. Results: When LDSC regression was applied to summary statistics obtained from meta-analysis, it led to an underestimation of heritability. This can occur if meta-analysis is used to combine studies of different ancestries leading to the caveat of the lack of an appropriate LD reference panel. Additionally meta-analyses often include studies with different phenotype definitions, that not only impacts heritability estimates but also makes them uninterpretable. Summary statistics generated from imputed variants, even those with high imputation accuracy, can lead to underestimation of heritability. For example, the heritability estimates for AD were reduced from 0.265 (se 0.148) to 0.160 (se 0.041) when imputed variants (INFO>0.9) were included compared to analyzing only genotype array variants. A decrease in heritability estimates was also observed when individual-level imputed variant data were analyzed using GCTA-GREML. Our findings highlight the caveats of estimating heritability using meta-analysis summary statistics or imputed data instead of genotyped or sequence data.

Background Physical activity levels are altered across neuropsychiatric disorders. While these traits are heritable, the genetic overlap between normal variation in activity levels and neuropsychiatric disorders that involve motor dysfunction such as schizophrenia and Parkinson's disease (PD) remains unexplored. Objectives To investigate the genetic overlap between physical activity, schizophrenia, and PD. Methods Multi-Trait Analysis of genome-wide association studies (GWAS) was used to boost the GWAS power for objectively measured physical activity (n=89,683) by leveraging three GWAS of self-reported activity (n=124,842-377,234). Genetic overlap between the activity, schizophrenia and PD was characterized using linkage disequilibrium score regression, causal mixture modeling, and local genetic correlations. Pleiotropic variants were identified using the conjunctional false discovery rate, annotated to genes, and investigated for enrichment of biological processes, tissue types and association with GWAS-catalog traits. Results Genetic correlations of physical activity with schizophrenia and PD were negligible (rg=-0.02-0.02, p>0.05), but polygenic overlap was substantial, reflecting mixed effect directions. We identified 32 independent variants shared with schizophrenia and 11 with PD, including CRHR1, MAPT and KANSL1 within the 17q21.31 region. Schizophrenia-shared variants mapped to genes differentially expressed in subcortical regions, especially amygdala and basal ganglia. Gene-set analyses revealed enrichment for mental health and cognitive-behavioural traits (schizophrenia-shared genes) versus structural brain phenotypes and neurodegenerative disorders (PD-shared genes). Conclusions Despite negligible genetic correlations, physical activity shares substantial genetic architecture with schizophrenia and PD. Shared genes implicated brain regions and traits spanning motor and cognitive-affective function, consistent with the psychomotor nature of physical activity.

BackgroundNon-invasive gamma entrainment using sensory stimulation (GENUS) is being investigated as a therapy for Alzheimer disease (AD), but the clinical course of participants who fail to show gamma center frequency plasticity remains unclear. We therefore examined whether cognitive decline observed during personalized GENUS was attenuated after cessation in participants stratified post hoc by CF change. MethodsThis case series was derived from an open-label proof-of-concept trial with extended follow-up (mean 26.3 months). Sixteen participants with amyloid-positive early AD completed 12 weeks of home-based daily flickering light stimulation (1 hour/day) at individualized gamma frequency, and 12 completed long-term follow-up. Participants were classified post hoc as ICF+ (CF increase [&ge;]2 Hz; n=5) or ICF- (no CF increase; n=7). MMSE trajectories from baseline to week 12 and from week 12 to final follow-up were analyzed using piecewise linear mixed-effects models. ResultsBaseline characteristics, including MMSE, did not differ significantly between groups. During intervention, MMSE was stable in ICF+ (+0.27 points/month; 95% CI, -0.20 to 0.73) but declined in ICF- (-0.69 points/month; 95% CI, -1.07 to -0.30; between-group p=0.011). After cessation, estimated slopes were similar in ICF+ and ICF- (-0.16 vs -0.17 points/month; p=0.968), and the phasexgroup interaction was significant (p=0.006). Medication intensification during follow-up was more common in ICF-, including antipsychotic initiation in 4 of 7 participants. ConclusionsIn this exploratory post hoc analysis, lack of CF plasticity was associated with accelerated cognitive decline during the intervention phase but not during follow-up. This temporal pattern is consistent with, but does not establish, a dechallenge-like safety signal. Given the small sample, post hoc stratification, and differential medication changes, these findings should be considered hypothesis-generating and require prospective validation with pre-defined electrophysiologic stratification. Trial registrationClinical Research Information Service (CRIS), Republic of Korea (KCT0010618); submitted 6 October 2022; first patient enrolled 2 February 2023. https://cris.nih.go.kr/cris/search/detailSearch.do?seq=31321.