Communications Psychology

Despite significant neurobiological and pathological overlaps, Alzheimers (AD) and Parkinsons (PD)--the primary threats to healthy aging--are still managed as distinct clinical entities. Standard machine learning exacerbates this fragmentation by prioritizing divergent markers over shared traits, obscuring the invariant foundations of neurodegeneration. This study introduces an explainable framework leveraging Importance Inversion Transfer (IIT) to identify candidate neuroanatomical features that show relative stability across both disorders. By prioritizing structural anchors invariant across the neurodegenerative spectrum, IIT isolates candidate shared structural features. Analysis of multi-regional brain volumes identifies eight shared anchors, confirmed via an inductive validation protocol with high diagnostic robustness (AUC = 0.894). Findings reveal a morphological continuum between healthy aging and neurodegeneration, suggesting the presence of partially shared structural substrates. These results are consistent with--though do not demonstrate--a potential common early-phase vulnerability across neurodegenerative conditions, as conceptualized by the Neurodegenerative Elderly Syndrome (NES) framework, establishing a possible paradigm for early, system-level diagnosis.

Large language models are increasingly used for clinical guidance while their parent companies introduce advertising. We tested whether pharmaceutical ads embedded in the prompts of 12 models from OpenAI, Anthropic, and Google shift drug recommendations across 258,660 API calls and four experiments probing distinct epistemic conditions. When two drugs were both guideline appropriate, advertising shifted selection of the advertised drug by +12.7 percentage points (P < 0.001), with some model scenario pairs shifting from 0% to 100%. Google models were the most susceptible (+29.8 pp), followed by OpenAI (+10.9 pp), while Anthropic models showed minimal change (+2.0 pp). When the advertised product lacked evidence or was clinically suboptimal, models resisted. This reveals a structured vulnerability: advertising does not override medical knowledge but fills the space where clinical evidence is underdetermined. An open response sub analysis (2,340 calls across three representative models) confirmed that advertising restructures free-text clinical reasoning: models echoed ad claims at 2.7 times the baseline rate while maintaining high stated confidence and rarely disclosing the ad. Susceptibility was provider dependent (Google: +29.8 pp; OpenAI: +10.9 pp; Anthropic: +2.0 pp). Because this bias operates within clinically correct answers, it is invisible to accuracy based evaluation, identifying a class of AI safety vulnerability that standard testing cannot detect.

Background: Large language models (LLMs) are increasingly used in mental health contexts, yet their detection of suicidal ideation is inconsistent, raising patient safety concerns. Objective: To evaluate whether an independent safety monitoring system improves detection of suicide risk compared with native LLM safeguards. Methods: We conducted a cross-sectional evaluation using 224 paired suicide-related clinical vignettes presented in a single-turn format under two conditions (with and without structured clinical information). Native LLM safeguard responses were compared with an independent supervisory safety architecture with asynchronous monitoring. The primary outcome was detection of suicide risk requiring intervention. Results: The supervisory system detected suicide risk in 205 of 224 evaluations (91.5%) versus 41 of 224 (18.3%) for native LLM safeguards. Among 168 discordant evaluations, 166 favored the supervisory system and 2 favored the LLM (matched odds ratio {approx}83.0). Both systems detected risk in 39 evaluations, and neither in 17. Detection was highest in scenarios with explicit suicidal ideation and lower in more ambiguous presentations. Conclusions: Native LLM safeguards frequently failed to detect suicide risk in this structured evaluation. An independent monitoring approach substantially improved detection, supporting the role of external safety systems in high-risk mental health applications of LLMs.

Current digital twin brain (DTB) models are typically optimized using single-modality functional neuroimaging data, which restricts their ability to simulate brain dynamics across multiple spatiotemporal scales. Here, we bridged this gap by developing a two-stage DTB (TS-DTB) modeling framework jointly constrained by functional magnetic resonance imaging (fMRI) and electroencephalography (EEG) data. Validation in both the healthy young and Alzheimers disease (AD) cohorts demonstrated that the TS-DTB model simultaneously captures subject-specific features of multi-scale brain dynamics. In particular, the TS-DTB models of AD patients successfully recapitulated spectral-temporal signatures of cognitive decline, mechanistically linking these deficits to excitation-inhibition (E-I) imbalances. By simulating responses to repetitive transcranial magnetic stimulation (rTMS), we further revealed that cognitive recovery in AD patients can be driven by E-I rebalancing via synaptic reconfiguration and background suppression. Overall, these findings underscore the potential of the TS-DTB to advance the mechanistic understanding of the brain and inform personalized digital therapeutics.

Dynamic touch requires the perceptual system to extract stable material properties from complex, evolving signals. We show that the tactile system relies on total spectral energy, the overall vibratory power of contact-induced transients, rather than waveform details or dominant frequency. Using a spectral energy compensation method, we conducted five psychophysical experiments in two degraded feedback scenarios: soft finger interfaces, where fingertip stiffness was reduced by an inflatable silicone bubble, and soft surface interactions, where participants tapped compliant foam surfaces. In both, participants reliably discriminated hardness and identified materials only when natural spectral energy profiles were preserved, independent of signal type. Judgments scaled systematically with energy level, and under conflicting cues, spectral energy dominated over frequency or compliance. These findings establish spectral energy as a governing cue in tactile perception, revealing a simple and robust computation akin to estimating mechanical work. This principle offers a generalizable framework for restoring touch in prosthetics, teleoperation, and immersive virtual environments. TeaserTotal spectral energy - not frequency - is the behaviorally relevant feature driving material perception through dynamic touch.

A central question in psycholinguistics in visual word recognition is whether morphologically complex words are obligatorily decomposed into stems and affixes during visual word recognition or whether whole-word access can occur when forms are frequent and familiar. The present study investigated how morphological complexity and lexical frequency jointly shape neural responses by leveraging Korean nominal inflection, whose transparent stem-suffix structure permits a clean dissociation between base (stem) frequency and surface (whole-word) frequency. Twenty-five native Korean speakers completed a rapid event-related fMRI lexical decision task involving simple and inflected nouns that varied parametrically in both frequency measures. Representational similarity analysis (RSA) revealed robust encoding of surface frequency--but not base frequency--in the inferior frontal gyrus (IFG) pars opercularis and supramarginal gyrus (SMG), with significantly stronger correlations for inflected than simple nouns. Univariate analyses converged with this result: surface frequency selectively increased activation for inflected nouns in inferior parietal regions, whereas base frequency showed no reliable effects in any ROI. These findings challenge models positing obligatory pre-lexical decomposition, instead supporting accounts in which morphological processing is shaped by post-lexical, usage-driven lexical statistics. Taken together, our findings shed light on a distributed perspective on morphological processing, suggesting that structural and statistical factors jointly constrain access to morphologically complex forms.

Path integration and spatial updating refer to the integration of self-motion information during navigation to update ones start location and the positions of other locations, respectively. Even though path integration has been described as a fundamental process whose output may serve as a building block for other navigational computations like spatial updating, the exact relationship between path integration and spatial updating is unknown. Here we addressed this question with an eye-tracking behavioral experiment and a subsequent fMRI study. Despite experiencing identical self-motion cues, participants showed differential eye fixation patterns and responded more quickly during spatial updating than during path integration, casting doubt on the fundamental role of path integration. Neuroimaging results showed that the precuneus and the dorsal premotor cortex were more activated during spatial updating, but the precuneus had stronger functional connectivity with the thalamus and the frontal cortex during path integration. Further supporting this dissociation, the two tasks invoked distinct brain-wide inter-regional functional networks. Together, the combined findings of both experiments suggest that spatial updating and path integration are dissociable navigation processes supported by distinct behavioral and neural mechanisms, rather than one process operating on the basis of the other.

Manuscript preparation is a critical bottleneck in scientific publishing, yet existing AI writing tools require cloud transmission of sensitive content, creating data-confidentiality barriers for clinical researchers. We introduce the Paper Analysis Tool (PAT), a free, multi-agent framework that deploys 31 specialized agents powered by small language models (SLMs) to audit manuscripts across multiple quality dimensions without external data transmission. Applied to three published clinical neurological papers, PAT generated 540 evaluable suggestions. Validation by two expert reviewers (R.B., A.G.) confirmed 391 actionable, high-value revisions (90% agreement), achieving a 72.4% overall usefulness accuracy spanning methodological, statistical, and visual domains. Furthermore, deterministic re-evaluation of 126 agent-suggested rewrite pairs using Phase 0 metrics confirmed text improvement: total word count decreased by 25%, passive voice prevalence dropped sharply from 35% to 5%, average sentence length decreased by 24%, long-sentence fraction fell by 67%, and the Flesch-Kincaid grade improved by 17% . Our validation confirms that systematic, agent-driven pre-submission review drives measurable improvements, successfully converting manuscript optimization from an opaque, manual endeavor into a transparent and rigorous scientific process. Manuscript preparation is a critical bottleneck in scientific publishing, yet existing AI writing tools require cloud transmission of sensitive content, creating data-confidentiality barriers for clinical researchers. We introduce the Paper Analysis Tool (PAT), a free, multi-agent framework that deploys 31 specialized agents powered by small language models (SLMs) to audit manuscripts across multiple quality dimensions without external data transmission. Applied to three published clinical neurological papers, PAT generated 540 evaluable suggestions. Independent validation by two expert reviewers (R.B., A.G.) confirmed 391 actionable, high-value revisions (90% agreement), achieving a 72.4% overall usefulness accuracy spanning methodological, statistical, and visual domains. Furthermore, deterministic re-evaluation of 126 suggested Phase 0 rewrite pairs confirmed text improvement: total word count decreased by 25%, passive voice prevalence dropped sharply from 35% to 5%, average sentence length decreased by 24%, and long-sentence fraction fell by 67%, and the Flesch-Kincaid grade improved modestly. Our validation confirms that systematic, agent-driven pre-submission review drives measurable improvements, successfully converting manuscript optimization from an opaque, manual endeavor into a transparent and rigorous scientific process.

Therapeutic brain stimulation is believed to target specific networks, but targeting approaches for memory remain debated. For other symptoms, neuromodulation targets have been localized by mapping connectivity of lesions and stimulation sites to specific symptoms. This approach has yielded networks for global memory, but it remains unclear whether it applies to specific types of memory. Here, we mapped connectivity of stimulation sites, lesions, and atrophy patterns associated with different memory types. We included 544 individuals across three datasets: transcranial magnetic stimulation (N=262), penetrating head trauma (N=169), and ischemic stroke (N=113). We identified a network preferentially connected to lesions and stimulation sites specifically associated with changes in visual memory. Of note, the direction of this effect was inverted depending on whether lesions or stimulation occurred at younger age or an older age, consistent with prior results. This age effect was replicated in an independent dataset of patients with preclinical Alzheimers disease (N=1240). To examine neuromodulation targets, we computed electrical field models for potential TMS sites that overlap with the networks derived from each stimulation or lesion dataset; the resulting targets intersected with established targets that demonstrated efficacy for treating memory impairment - precuneus, cortical-hippocampal network, and dorsolateral prefrontal cortex - with peak intersection at medial posterior parietal lobe, angular gyrus, and left anterior middle frontal gyrus, respectively. Future head-to-head clinical trials are needed to systematically compare these proposed neuromodulation targets against each other. One Sentence SummaryNeuromodulation targets for visual memory diverge by age at the time of injury or stimulation.

Introduction: The Nationwide Quality of Life Scale (NQLS) is a brief, mental-health focused quality of life (QoL) scale with seven items that are non-overlapping with symptom scales. We developed a parent version (P-NQLS), obtained national norms, and calculated psychometric properties for the P-NQLS. Methods: Parents (N=2251) of children aged 6-18 years who were representative of the U.S. population on key demographics completed the P-NQLS along with measures of depression, suicidality, internalizing, externalizing, and attention symptoms. We assessed the P-NQLS's factor structure through exploratory factor analysis (EFA) and evaluated its internal reliability and convergent validity. Age- and sex-specific norms were established using GAMLSS with BCPE distributions and P-spline smoothers, with percentile curves and tables (5th-95th) provided. Results: EFA suggested a one-factor solution for P-NQLS in the national sample. The scale showed good internal consistency (Cronbach's alpha=0.85). P-NQLS total scores (M=20.7, SD=4.7, range=0-28, higher scores indicate higher QoL) were negatively correlated (all p<.0001) with depression (Pearson's r=-0.47), suicidality (r=-0.50), internalizing (r=-0.43), externalizing (r=-0.41), and attention (r=-0.37) symptoms. P-NQLS scores declined steadily with age in both sexes, with the most pronounced decreases (3-5 points) observed at lower percentiles (5th, 10th), suggesting greater age-related decline among children with lower baselines. Females scored slightly higher than males across most ages and percentile levels, though the differences were within one point. Conclusions: The newly created P-NQLS, a 7-item parent-reported QoL scale with one underlying factor, demonstrates strong reliability and validity and has robust national norms for youth aged 6-18.

Alzheimers disease (AD) is characterized by diminished capacity to mount adaptive cellular stress responses required to maintain energy homeostasis and proteostasis. An emerging therapeutic strategy is to restore adaptive stress responses by inducing mild energetic stress through inhibition of mitochondrial complex I (mtCI). However, pharmacological inhibition of the respiratory chain has remained challenging, as it can induce bioenergetic failure rather than beneficial signaling. Here, we describe C273, a brain-penetrant small molecule that delivers controlled, weak attenuation of mtCI activity to therapeutically restore endogenous adaptive stress pathways. This work establishes a first-in-class mechanism in which calibrated activation of multifaceted adaptive mechanisms enhances cellular resilience, rather than impairing mitochondrial function. Structure-activity relationship optimization yielded a compound with high potency against A{beta}-induced cellular toxicity, strong selectivity for mtCI, and favorable drug-like properties. C273 demonstrated excellent oral bioavailability, metabolic stability in mouse, rat, and human microsomes, minimal CYP liabilities, and a clean ancillary pharmacology profile in the Eurofins CEREP44 panel. In vivo, C273 readily crosses the blood-brain barrier and activates AMP-activated protein kinase (AMPK), initiating a coordinated hormetic response characterized by enhanced antioxidant defenses, suppression of inflammatory signaling, induction of autophagy, and increased mitochondrial biogenesis and turnover. Genetic deletion of AMPK1/2 abolished these responses, establishing AMPK as a critical mediator of C273 activity. Pharmacological competition experiments further confirmed the target, as pretreatment with non-toxic concentrations of rotenone blocked C273 interaction with the quinone-binding site of mtCI and eliminated its neuroprotective effects. Repeated oral administration of C273 (20-80 mg/kg/day) to wild-type mice for one month produced no detectable cardiac or hepatic toxicity, indicating a favorable in vivo safety margin. Importantly, C273 activated these mechanisms and reduced A{beta} and p-Tau levels in induced pluripotent stem cell-derived cerebral organoids from patients with sporadic AD. Collectively, these results establish controlled mtCI modulation as a therapeutic strategy and position C273 as a promising disease-modifying candidate for AD.

Importance: Suicide is a leading cause of death in the United States with risk strongly influenced by Interpersonal trauma, contributing to treatment resistance and clinical complexity. Objective: To assess clinical and genetic factors in individuals who died from suicide, with and without interpersonal trauma exposure. Design: Individuals who died from suicide with and without trauma were compared in a retrospective case-case design. Prevalence of 19 broad clinical categories was assessed between groups. Results directed selection of 42 clinical subcategories, and 40 polygenic scores (PGS) for further assessment. Multivariable logistic regression models, adjusted for critical covariates and multiple tests, were formulated. Models were also stratified by age group (<26yo and >=26yo), sex, and age/sex. Setting: A population-based evaluation of comorbidity and polygenic scoring in two suicide death subgroups. Participants: A total of 8 738 Utah Suicide Mortality Research Study individuals (23.9% female, average age = 42.6 yo) who died from suicide were evaluated, divided into trauma (N = 1 091) and non-trauma exposed (N = 7 647) individuals. A subset of unrelated European genotyped individuals was also assessed in PGS analyses (Trauma N = 491; Non-trauma N = 3 233). Exposures: Trauma is here defined as interpersonal trauma exposure, including abuse, assault, and neglect from International Classification of Disease coding. Main Outcomes and Measures: Prevalence of comorbid clinical sub/categories and PGS enrichment in trauma versus non-trauma exposed suicide deaths. Results: Overall, trauma-exposed individuals died from suicide earlier (mean age of 38.1 yo versus 43.3 yo; P <0.0001) and were disproportionately female (38% versus 21%, OR = 3.3, CI = 2.9-3.8). Prevalence of asphyxiation and overdose methods, prior suicidality, psychiatric diagnoses, and substance use (OR range = 1.3-3.7) were elevated in trauma exposed individuals who died from suicide. Genetic PGS were also elevated in trauma-exposed individuals who died from suicide for depression, bipolar disorder, cannabis use, PTSD, insomnia, and schizophrenia (OR range = 1.1-1.4) with ADHD and opioid use showing uniquely elevated PGS in trauma exposed males (OR range = 1.2-1.4). Conclusions and Relevance: Results demonstrated multiple convergent lines of age- and sex-specific evidence differentiating trauma-exposed from non-trauma exposed suicide death. Such findings suggest unique biological backgrounds and may refine identification and treatment of this high-risk group.

Dementia incidence has been declining in Western societies for decades, but whether this reflects higher cognitive capacity entering old age, slower cognitive decline, or both remains unresolved. Analysing ~783,000 episodic memory assessments from ~219,000 individuals across five longitudinal cohorts, we find that later-born cohorts benefit from a double dividend: higher memory levels entering old age and slower rates of decline. The projected 20-year cohort advantage at age 80 is of sufficient magnitude to plausibly account for the observed 13% per-decade decline in dementia incidence reported in meta-analyses. Generational gains are disproportionately concentrated among the fastest-declining individuals, and are reflected in lower hippocampal atrophy rates in an independent sample. A formal bounding analysis shows that the double dividend is robust across a range of plausible period assumptions, consistent with environmental conditions operating across the lifespan having reshaped the architecture of human cognitive aging.

AI documentation frameworks remain poorly designed for point-of-care use, leaving clinicians without actionable information on how to use clinical AI models when they need it most. We developed the Clinician Model Card, an interactive, clinician-centered documentation tool, and evaluated it in a sequential exploratory mixed-methods study: interviews with 12 physicians informed iterative co-design, evaluated in a national survey of 129 physicians across Germany. The tool was well-received: 84% agreed it should be routinely available, and 66% considered its content relevant to clinical decision-making. Yet comprehensibility of statistical performance metrics remained poor despite targeted interventions: only 32% understood the Validation & Performance section well, and fewer than 54% correctly interpreted AUROC or PPV, with AI literacy as strong predictor of comprehension. We propose empirically derived design principles for clinician-centered AI documentation. Effective AI transparency requires not only clinician-friendly design and workflow integration, but sustained investment in AI literacy.

Automating Hierarchical Condition Category (HCC) assignment directly from unstructured electronic health record (EHR) notes remains an important but understudied problem in clinical informatics. We present HCC-Coder, an end to end NLP system that maps narrative documentation to 115 Centers for Medicare & Medicaid Services(CMS) HCC codes in a multi-label setting. On the test dataset, HCC-Coder achieves a macro-F1 of 0.779 and a micro-F1 of 0.756, with a macro-sensitivity of 0.819 and macro-specificity of 0.998. By contrast, Generative Pre-trained Transformer (GPT)-4o achieves highest score of a macro-F1 of 0.735 and a micro-F1 of 0.708 under five-shot prompting. The fine-tuned model demonstrates consistent absolute improvements of 4%-5% in F1-scores over GPT-4o. To address severe label imbalance, we incorporate inverse-frequency weighting and per-label threshold calibration. These findings suggest that domain-adapted transformers provide more balanced and reliable performance than prompt-based large language models for hierarchical clinical coding and risk adjustment.

Background. Federated learning (FL) enables collaborative model training across institutions without sharing patient-level data. However, standard FL algorithms such as FedAvg degrade under non-independently and non-identically distributed (non-IID) data, a prevalent condition when patient demographics, scanner hardware, and disease prevalence differ across hospital sites. Objective. We propose iPS-MFFL (Individualized Per-Site Meta-Federated Feature Learning), a federated framework with a hierarchical local-model architecture that addresses non-IID heterogeneity through (1) a shared feature extractor, (2) multiple weak-learner classification heads that can be trained with heterogeneous training objectives to promote complementary decision boundaries, (3) independent per-learner server aggregation so that each weak learner's parameters are averaged only with its counterparts at other clients, and (4) a lightweight meta-model, itself federated, that adaptively stacks the weak-learner outputs. Methods. We evaluate on the Brain Tumor MRI Classification dataset (7,200 images; 4 classes: glioma, meningioma, pituitary tumor, no tumor) partitioned across K = 5 simulated hospital sites using Dirichlet non-IID sampling (alpha = 0.3). Four baselines are compared: Local-only training, FedAvg, FedProx, and Freeze-FT. All experiments are repeated over three random seeds (13, 42, 2025) and evaluated using paired t-tests, Cohen's d effect sizes, and post-hoc power analysis.

In this study, we investigate gender bias in a Retrieval-Augmented Generation (RAG) based AI assistant developed for Finnish wellbeing services counties. We tested the system using 36 clinically relevant queries, each rendered in three gendered variants (male, female, gender-neutral), and evaluated responses using both an LLM-as-a-judge approach and a human expert panel consisting of a physician and a sociologist specializing in ethics. We observed substantial and clinically significant differences across gendered variants, including differential treatment urgency, inappropriate symptom associations, and misidentification of clinical context. Female variants disproportionately framed responses around childcare and reproductive health regardless of clinical relevance, reflecting societal stereotypes rather than medical reasoning. Bias manifested both at the LLM generation stage and the RAG retrieval stage, in several cases causing the model to hallucinate responses entirely. Some bias patterns were persistent across repeated runs, while others appeared inconsistently, highlighting the challenge of distinguishing systematic bias from stochastic variation.

Sensory systems must represent a vast range of stimulus dimensions and energy whilst subject to metabolic constraints. Efficient-coding theory predicts that neural adaptation re-allocates a relatively limited range of neural activity toward the most informative stimulus values, but it is unclear how subtle peripheral lesions shift this operating point in central circuits. Hearing is a stringent test because sound level varies enormously across environments, yet clinical assessment still relies heavily on tone-detection thresholds that can miss listening deficits in noise. We analyzed extracellular recordings from single neurons in the gerbil auditory midbrain across 14 animals in four experimental groups exposed to unfolding distributions of sound intensities drawn either uniformly from a wide range (24-96 decibels) of sound pressure levels or from contexts in which 80% of levels were restricted to a 12-decibel high-probability range. For each context we summarized each neurons rate-intensity input-output function by an effective threshold and gain, and we interpreted the resulting threshold-gain distributions with an information-cost model that trades bits of stimulus information against a penalty on mean spiking. Noise exposure consistent with loss of synapses between inner-ear cells and auditory nerve fibers altered gain modulation across acoustic contexts, with noise-exposed animals showing compressed gain adjustments relative to controls; within the information-cost framework, the clearest hidden-hearing-loss effect was a quiet-context utility advantage concentrated in the low-threshold neural population, whereas moderate-to-loud contexts showed weaker or absent group differences. Temporary conductive attenuation caused by ear-canal plugging shifted effective thresholds to higher sound levels, with incomplete recovery after plug removal; the corresponding optimization-prior trajectories were consistent with incomplete rapid renormalization but were weaker than the hidden-hearing-loss effect. These results support an efficient-coding interpretation of altered central auditory representations after subtle lesions and provide a quantitative, context-based framework for comparing mechanisms of hearing difficulty beyond threshold-only tests and Fisher information alone. Author SummaryEveryday hearing is an ecological challenge for the auditory system: we must follow speech while background sounds fluctuate and overlap. Standard tests emphasize tone-detection thresholds, but many listeners struggle in noise even when thresholds appear normal. We asked whether subtle peripheral changes shift how the auditory brain trades information for neural effort. We analyzed recordings from single neurons in the gerbil auditory midbrain during sound environments with different loudness statistics, including ones dominated by a narrow intensity range. Using information-theoretic measures, we quantified how much spikes distinguished sound-level categories and related this to the amount of spiking produced. Noise exposure consistent with inner-ear synaptic loss altered gain modulation across acoustic contexts and most strongly improved model-based coding utility in quieter settings, but reduced adaptation and efficiency as sound environments became louder. Temporary ear-canal plugging raised effective response thresholds substantially above both control and synaptopathy groups, with only partial recovery immediately after plug removal. By mapping both manipulations onto a common information-versus-cost scale, we highlight context-dependent metrics that may prove more informative than threshold audiograms for subtle hearing problems.

Menarche is a normative milestone of female puberty, yet its role in adolescent mental health and brain development remains poorly understood. Using longitudinal data from 5,016 females (7 annual visits, ages 10-16 years) in the Adolescent Brain Cognitive Development Study, we found that menarche onset functions as an inflection point for the development of internalizing symptoms and gross brain morphometry. The onset of menarche, largely independent of timing and socio-environmental factors, preceded a significant spike in internalizing symptoms, while altering the rate of ongoing structural brain development. Following menarche onset, individuals with faster declines in gray matter volume and surface area also had heightened internalizing symptoms. These findings suggest that menarche is not only a reproductive milestone but a neuroendocrine driver of adolescent brain and mental health trajectories. This normative and easily identifiable marker could define a critical window for mental health screenings with greater precision than current age-based guidelines.

Age-related hearing loss is linked to loneliness and poorer cognitive health, but it remains unclear whether loneliness helps explain associations between hearing difficulties and cognitive performance or dementia, and whether these patterns reflect causal pathways or shared underlying liability. In this preregistered study, we triangulated analyses across multiple data sources spanning approximately 18 years of observational data with 8 sources of molecular genetic information to examine whether loneliness helps explain the association between hearing difficulty and cognitive performance, Alzheimer's disease dementia, and all-cause dementia, and whether hearing-aid use may buffer this association. In longitudinal parallel-process latent growth curve models (N = 10,375) using nine waves of longitudinal data from the Survey of Health, Ageing and Retirement in Europe (SHARE), poorer hearing was associated with greater loneliness, and greater loneliness was associated with poorer cognitive performance, consistent with partial mediation. In contrast, worsening hearing over time was not clearly associated with increasing loneliness over time. Cumulative hearing-aid use did not appear to alter long-term loneliness trajectories, although current hearing-aid use weakened the concurrent association between poorer hearing and greater loneliness. In genetic analyses, we found little evidence that hearing phenotypes or loneliness had clear total or indirect effects on Alzheimer's disease dementia or all-cause dementia. Analyses accounting for shared genetic liability with neuroticism provided some evidence linking loneliness with poorer cognitive performance, and colocalization analyses further suggested shared genetic architecture across hearing, loneliness, cognition, and neuroticism-related traits. Overall, the findings support a robust cross-domain association between poorer hearing, greater loneliness, and poorer cognitive performance, while suggesting that long-term change and genetic evidence are more consistent with shared liability than with a single causal pathway.