JMIR mHealth and uHealth

Background Wearable-based studies have largely examined activity and sleep using static summaries or single time windows, potentially missing how chronic patterns and recent behavioral changes jointly relate to depressive symptom severity. We evaluated whether combining long-term habitual behavior with short-term dynamics improves characterization of moderate-to-severe depressive symptoms. Methods We analyzed Fitbit data from All of Us participants with Patient Health Questionnaire-9 (PHQ-9) assessments, defining moderate-to-severe symptoms as PHQ-9 [≥] 10 (N=248). Logistic regression evaluated long-term measures (past-year step count and awake time after sleep onset) and short-term dynamics (30-day step decline and 30-day sleep duration variability), adjusting for demographics. Performance was assessed via repeated stratified 10-fold cross-validation. Results Thirty percent of participants (n = 74) had moderate-to-severe depressive symptoms. Higher long-term step count was associated with lower odds of elevated symptoms (OR = 0.75 per 1,000 steps/day), greater awake time after sleep onset with higher odds (OR = 1.27 per 1%), a 30-day step decline with higher odds (OR = 2.70), and greater 30-day sleep variability with higher odds (OR = 1.07 per percentage point). Short-term dynamics provided complementary information beyond long-term measures alone. The combined model achieved the highest discrimination (area under the curve [AUC] = 0.80 vs. 0.73 demographics-only), though findings should be interpreted as exploratory given the modest sample size. Limitations The sample was modest in size (N = 248), PHQ-9 reflects symptom severity rather than clinical diagnosis, causal inference is not possible given the cross-sectional outcome assessment, and Fitbit users may not represent broader populations. Conclusions Long-term behavioral patterns and short-term changes in activity and sleep were associated with depressive symptom severity, supporting wearable-derived measures as potential adjunctive markers in mental health research.

Abstract Importance: Irregular sleep-wake patterns have been associated with poor health and cognitive outcomes, yet evidence linking 24-hour sleep-wake regularity to cognitive decline or dementia remains inconsistent. Particularly, regularity can be measured as regularity of rest-wake, sleep-wake or overall 24-hour activity, but it is unclear which aspects are most relevant for cognitive aging. Objective: To assess associations of rest-wake, sleep-wake, and 24-hour activity regularity with cognitive decline and dementia risk. Design: Observational prospective study comprised of six US and European cohorts: MrOS (sleep study between 2003-2005, mean follow-up: 7.1 years), Rotterdam Study (2004-2007, 11.6 years), MESA (2010-2013, 8.2 years), MAP (2005-2018, 7.2 years), Whitehall II (2012-2013, 6.9 years), and UKB (2013-2015, 7.9 years). Setting: Cohort-specific estimates were pooled using random-effects meta-analysis. Analyses were done between June 2025 and March 2026. Participants 74,733 dementia-free adults with multi-day actigraphy were included across cohorts: MrOS (age: 67-96 years, female:0%), MESA (54-95y, female:54.6%), Rotterdam Study (46-98y, female:55.0%), MAP (56-100y, female:77.1%), Whitehall II (59-83y, female:25.9%), and UKB (55-78y, female:55.5%). Exposure: Day-to-day rest-wake regularity (Rest Regularity Index, RRI), day-to-day sleep-wake regularity (Sleep Regularity Index, SRI), and 24-hour activity regularity (Interdaily Stability, IS) were derived from multi-day actigraphy. Main Outcome: Outcomes were risk of dementia and changes in global cognition. Results: Across six cohorts, 1,906 dementia cases occurred among 74,733 participants. After adjusting for demographics, health behaviors, depressive symptoms and cardiovascular comorbidities, each 1-SD higher regularity score was associated with an 9-14% lower dementia risk (pooled hazard ratios: RRI 0.86 95%CI: [0.79-0.95]; SRI 0.87[0.79-0.97]; IS: 0.91[0.88-0.95]). Associations were approximately linear. Age-stratified analyses showed directionally stronger associations among adults aged < 65, although meta-regression did not support an interaction(p > 0.55). Greater regularity was associated with modestly slower decline in global cognition (pooled {beta} per 1-SD higher score of RRI per year: 0.003, 95%CI [0.001-0.006]). Conclusions & Relevance: Greater regularity of rest-wake, sleep-wake, and 24-hour activity rhythms was associated with lower dementia risk and modestly slower global cognitive decline. These findings suggest that 24-hour sleep-wake regularity is a relevant behavioral marker of cognitive aging and may inform future efforts to identify or intervene on early risk.

Introduction: Actigraphy sleep-wake classification methods increasingly seek to leverage raw acceleration data and machine-learning-based classification, but performance evaluation in pediatrics is limited. We trained machine-learning models using pediatric data and compared their sleep-wake classification performance with existing algorithms for children. Methods: Sixty-five children (46% female, ages 5.3 to 17.7 years) completed in-lab overnight polysomnography and wore a GENEActiv device on their non-dominant wrist. The acceleration data were converted into 30-second epochs and aligned with physician-scored sleep-wake data from electroencephalography. Seven machine-learning models were trained using leave-one-subject-out cross-validation. Epoch-by-epoch analyses generated performance metrics (e.g., balanced accuracy [BA]) and discrepancy analyses provided overall sleep duration bias estimates. The combination of highest performance and least bias was used to rank using Euclidean distance scores - where a lower score represents closer to perfect performance and zero bias. For benchmarking, we included GGIR sleep scoring algorithms and an adult trained random forest classifier. Results: Overall, 560.1 hours of polysomnography and actigraphy data were collected (74.4% of epochs were scored as sleep). The pediatric-trained local-global long-short term memory (LSTM) classifier had the most optimal epoch-by-epoch performance (e.g., BA=0.85, sensitivity=0.88, specificity=0.83, ROC-AUC=0.95, and Cohen kappa=0.67). These metrics exceeded that of an adult-trained random forest classifier and GGIR-based algorithms. Discrepancy analyses revealed that overall sleep duration was underestimated by an average of 25 minutes using the LSTM classifier with no proportional bias. Conclusion: We trained seven pediatric sleep-wake classifiers that had strong ability to detect sleep and wake, with the LSTM classifier being most optimal.

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.

The Asian population in Los Angeles is among the largest and most heterogeneous in the U.S. This is true culturally and health-wise. Older Asians have differing risks for cardiovascular and cardiometabolic disease, depending on their ethnicity, health literacy, and lifestyle choices. This pilot examines several of these factors in a small but diverse group of older Asian adults who attended community health events from 2024-2025. Self-reported and biometric data were collected at five such events hosted by the Asian Pacific Health Corps at UCLA. The pilot generated health literacy and lifestyle (HLL) scores for all participating attendees and explored how they relate to their socio-demographics, healthcare habits, and predictions of their own health data. Overall, there were significantly more females than males with higher HLL scores (p = 0.027). College education (p = 0.028) and "normal" ranges for biometric data (e.g., blood pressure, BMI, blood glucose, cholesterol) were related to higher median HLL scores. With a few exceptions, fewer than 50% accurately predicted their biometric numbers regardless of HLL scores, suggesting a disconnect between perception and reality, and that better provider-patient communication may help foster greater patient understanding about their chronic conditions. These HLL score distributions indicate that educational attainment, better awareness of one's health, and high health literacy are individual factors that may influence older Asians' understanding and potential approach to managing their health conditions.

Background: Wearable technologies enable scalable and continuous monitoring of emotional states through passive sensing of physiological and behavioral signals. However, conventional learning approaches often struggle to model the complex temporal, contextual, and relational dependencies underlying human emotions. To address these limitations, we propose a graph-based framework that represents multimodal wearable observations as heterogeneous knowledge graphs enriched with semantic information derived from Large Language Models (LLMs), enabling richer contextual understanding beyond raw sensor measurements. Methods: We constructed a heterogeneous knowledge graph using multimodal Fitbit physiological signals and affective self-report data collected from 45 users. Framing mood prediction and emotion detection was formulated as both binary and ternary node classification tasks. We evaluated five baseline heterogeneous Graph Neural Network (GNN) architectures and compared them with the proposed Semantically Gated Augmented Graph Neural Network (SeGA-GNN) framework, which dynamically integrates LLM-generated semantic embeddings into graph representations through a gated cross-modal fusion mechanism. Results: The baseline GNN models achieved strong performance, with classification accuracies ranging from 0.7525 to 0.9739 for binary classification and 0.6249 to 0.9699 for ternary classification. The proposed SeGA framework consistently improved predictive performance across most architectures. In particular, semantic augmentation transformed the HAN model from moderate baseline performance into near-perfect emotion recognition capability, achieving SeGA-HAN Accuracy = 0.9988 and AUC = 1.0000 for binary classification and Accuracy = 0.9979 and AUC = 1.0000 for ternary classification. Discussion and Conclusion: Integrating LLM-derived semantic contextualization into heterogeneous graph learning enables effective modeling of contextual information that is not directly captured by wearable physiological signals alone. The proposed SeGA-GNN framework demonstrates that adaptive semantic fusion substantially improves the accuracy, robustness, and interpretability of wearable-based emotion detection. These findings establish a promising direction for next-generation wearable affective computing systems and intelligent emotion-aware applications.

Background: Conventional psychiatric screening instruments summarize symptoms within individual scales and prioritize cases with high single-instrument additive score severity. This design treats items as independent within instruments and ignores cross-instrument covariance structure, making it insensitive to respondents whose responses are distributed across multiple domains in unusual combinations that remain below threshold on every individual scale. Methods: We analyzed two cohorts spanning older and younger adults. Item prompts from depression, stress, anxiety, and sleep instruments were embedded into a shared semantic space using a pretrained sentence encoder. Principal component analysis of the item-prompt embeddings alone---with no use of respondent data at this stage---was used to construct a low-dimensional subspace retaining 80\% of variance in the item embedding matrix. Normalized participant responses were then projected into this subspace, with Jaccard-based stability analysis used as a check on dimensional robustness. Multivariate deviation from the cohort norm was quantified with Mahalanobis distance using Ledoit-Wolf covariance regularization. Candidate outliers were defined by the empirical 95th percentile of the cohort-specific distance distribution. To isolate response configurations not already captured by conventional single-instrument extreme-value logic, we excluded all outlier respondents who had endorsed any individual item at the maximum value of its Likert scale on any instrument. For the remaining outliers, anomalous components were backtracked to their original item loadings for interpretation. Results: In the older-adult Health and Retirement Study (HRS) cohort, principal component analysis of 27 item-prompt embeddings showed that a 10-dimensional subspace provided a stable representation of cross-instrument semantic structure. In the younger-adult Xinxiang cohort the corresponding stable solution was 16-dimensional. In each cohort, seven respondents remained as multivariate outliers despite falling below every single-instrument extreme-value threshold. These cases were not characterized by uniformly severe symptom scores but by unusual cross-domain response configurations that became visible only in the shared semantic covariance subspace. The response structure of the retained configurations differed across cohorts: older-adult cases more often involved weak endorsement of mood-labeled items alongside nonzero body- and sleep-related responses, whereas younger-adult cases more often involved incomplete response configurations spanning mood, sleep, stress, and self-harm-related items. Conclusions: A semantically aligned, auditable covariance subspace provides a practical tool for flagging unusual multivariate response configurations that single-instrument additive screening may not flag. The method is interpretable at the level of original item contributions. It should be understood as a hypothesis-generating screen for unusual response configurations requiring further clinical assessment, not as a diagnostic instrument. Outcome validity remains to be established by prospective study.

Background: Circadian rest-activity rhythms weaken with age, but whether sleep disorders modify this trajectory is unknown. Methods: We analyzed wrist accelerometry data from 4,386 participants aged 6-80 years in the 2011-2012 National Health and Nutrition Examination Survey (NHANES). Circadian features were extracted using cosinor analysis and nonparametric methods; a Circadian Disruption Index (CDI) was constructed from five standardized components. Survey-weighted regression with natural cubic splines and Wald F-tests tested age-by-sleep-disorder interactions using Taylor series linearization for variance estimation. Results: Doctor-diagnosed sleep disorder (N = 360, 8.2%) was associated with significantly different age-related trajectories of amplitude (F(2,17) = 11.24, p = 0.0008) and MESOR (F(2,17) = 8.22, p = 0.0032), both surviving Bonferroni correction (p < 0.006). CDI was higher in those with a sleep disorder (0.290 vs. 0.131, p < 0.001) and was independently associated with higher BMI (beta = 1.33 kg/m2, p < 0.001), higher HbA1c (beta = 0.089%, p = 0.004), greater diabetes prevalence (beta = 3.8 percentage points, p < 0.001), and worse depressive symptoms (beta = 0.43 PHQ-9 points, p = 0.020). Sensitivity analyses using a broader sleep problem exposure did not replicate these interactions. Conclusions: Doctor-diagnosed sleep disorders are associated with an altered age-related decline in circadian amplitude and mean activity level. CDI was independently linked to cardiometabolic and depressive outcomes, supporting a mechanistic connection between clinically significant sleep pathology and circadian disruption across the lifespan.

Background. Complex cases in specialized pediatric care require consistent adherence to evidence-based clinical pathways and protocols to ensure safe, high-quality, and equitable care. Currently, clinical pathways and supporting documentation are frequently distributed across multiple platforms, leading to fragmentation. Human-centered design principles can guide the development of healthcare technologies that minimize cognitive load and support rapid, efficient access to relevant information in clinical settings. The purpose of this study is to design and evaluate perceived usability of a pediatric cardiac center digital guideline management system that is embedded within the electronic health record leveraging human-centered design. Methods. This study used a mixed-methods usability evaluation to assess a digital guideline management system prototype embedded into clinical workflow. Through human-centered design principles, the prototype provides a centralized digital document library that organizes cardiac-specific clinical pathways, guidelines, procedures, and related resources. A small but diverse sample, encompassing a wide variety of roles and clinical areas within the pediatric cardiac center, was recruited to evaluate the perceived usability of the prototype. Usability was evaluated by stakeholders using the validated System Usability Scale (SUS) with additional optional questions to understand perceptions of the information architecture and clinical value. Results. Preliminary usability testing showed a mean SUS composite score of 76.5, indicating above average usability. Questions related to the complexity of the system and user confidence received high scores across participants. Lower scores were observed for questions related to usage frequency and ability to learn the system very quickly. Conclusion. Leveraging human-centered design when building a digital guideline management system embedded within clinical workflow revealed positive perception from participants. By centralizing access to clinical resources, this prototype can reduce current-state fragmentation. Further evaluation of larger samples is needed to develop a list of future recommendations.

Pharmacogenomics (PGx) offers a pathway towards personalised medicine, which relies on health consumer involvement in making informed decisions. As consumers increasingly seek health information online, high-quality digital resources are essential to support informed consent and shared decision making. The complexity of PGx and widespread limitations in health literacy raise concerns about whether existing consumer-facing online PGx resources are understandable and sufficiently comprehensive. This study evaluates the readability, visual design, and informational quality of publicly available online written PGx health information. Twenty-three webpages met inclusion criteria. The mean readability corresponded to approximately 15 years of formal education (university level), substantially exceeding the Australian Government's recommended Year 7 reading level for public health materials. Informational quality was generally low, with most webpages being rated as poor or very poor. In contrast, visual design quality was relatively strong, with webpages achieving on average around three-quarters of the criteria. Although the visual presentation of PGx webpages is generally professional, their high reading difficulty and limited discussion of treatment choices and uncertainties reduce their usefulness for health consumer education. Improving readability, clearly communicating risks and limitations, and incorporating decision-support features may enhance the ability of online resources to support informed consent and shared decision making.

Abstract Aims To examine which demographic groups nicotine pouch advertisers chose to target on social media, and which groups Meta's algorithms actually delivered the adverts to. Design Cross-sectional analysis of advert-level data from the Meta Ad Library. Setting Meta social media platforms (including Facebook and Instagram) in the UK. Cases A random sample of 741 nicotine pouch adverts shown in the 12 months up to December 2025, and a comparison sample of 1,125 general adverts. Analyses of reach were restricted to adverts eligible for all genders and adult ages (444 pouch adverts; 674 general). Measurements Outcomes were advertiser-set gender and age-group targeting criteria (i.e., groups eligible to be shown each advert) and estimated advert reach to each group (i.e., number of people who saw each advert). Male-to-female reach ratios within age groups, and reach ratios comparing age groups, were calculated per advert and summarised using geometric means. To assess whether patterns were pouch-specific, comparisons with general adverts were made using ratios of reach ratios (RRR). Findings Advertisers of nicotine pouches targeted a broad sample; most adverts (79.1%; 586/741) were eligible to be shown to all genders, the remainder were restricted to men only. All were restricted to adults (minimum age 18 years) and most (95.6%; 708/741) had no upper age limit. Despite this, of pouch adverts eligible to be shown to all adults, adverts were more likely to reach men, particularly among younger men. Among 18-24-year-olds, pouch adverts reached around ten times as many men as women (RR 10.0, 95% CI 8.7-11.5), compared with a slight skew towards women for general adverts (RR 0.81, 95% CI 0.71-0.94), corresponding to an RRR of 12.3 (95% CI 10.0-15.1). Pouch adverts also showed a skew in reach towards younger age groups. Relative to those aged 35-44 years, reach was higher among 18-24-year-olds for nicotine pouch adverts (RR 1.33, 95% CI 1.17-1.51) but much lower for general adverts (RR 0.19, 95% CI 0.17-0.21), corresponding to an RRR of 7.0 (95% CI 6.0-8.2). Conclusions Nicotine pouch adverts on social media are often eligible to be shown broadly to all demographic groups but are disproportionately delivered to young men.

Background: Individuals with mild cognitive impairment (MCI) experience cognitive and functional declines that can negatively impact mood and reduce feelings of self-efficacy. These changes can also lead to elevated distress in care partners (CPs). Therefore, interventions that address quality of life and psychosocial factors in people with MCI and their CPs are needed. Objective: The present study evaluated the impact of a multidomain lifestyle program, the Cognitive Empowerment Program (CEP), on changes in psychosocial functioning, particularly empowerment, in people with MCI and their CPs. Methods: Participants were 94 people with MCI (Mean= 75.1 years old, 45.7% female, 81.9% white) and their CPs (Mean= 69.1 years old, 71.3% female, 87.3% white) that completed the 12-month CEP program comprised of physical, cognitive, and psychosocial interventions. Questionnaires were administered pre- and post-program to assess empowerment, self-efficacy, meaning and purpose, depression, and stress in participants with MCI alongside empowerment, depression, stress, and caregiving burden in CPs. Results: After completing the CEP program, participants with MCI endorsed higher empowerment and self-efficacy as well as fewer symptoms of depression and perceived stress. CPs endorsed feeling more empowered despite elevated caregiver burden. Conclusions: These results suggest multidomain lifestyle programs can positively impact wellbeing in MCI. Future research should focus on refining delivery models, exploring integration with pharmacological treatments, prioritizing inclusion of diverse populations, and measuring long-term outcomes to strengthen the reach and impact of programs like CEP.

Aims. The risk of cognitive decline after losing a spouse remained mixed. This study aims to investigate the association between spousal loss and risk of cognitive decline, assess whether this association varies by sex and age, and identify modifiable factors. Methods. We conducted a prospective cohort study using harmonized data from six population-based aging surveys: the US Health and Retirement Study and its sister surveys in England, Mexico, China, India, and South Africa, incorporating their respective Harmonized Cognitive Assessment Protocol (HCAP) sub-studies. Spousal loss (yes vs no) was the exposure. Cognitive outcomes (i.e., orientation, memory, executive function, and language), were assessed using HCAP neuropsychological batteries. We conducted parallel analyses in six cohorts. Associations between spousal loss and cognitive outcomes were estimated using generalized linear models, and summarised estimates were derived via random-effects meta-analyses. Sex stratification and restricted cubic spines were used to examine how these associations vary by sex and age, respectively. Results. The analytical cohort consisted of 18,551 individuals aged 61.22 (SD 6.30) to 71.37 (SD 7.33) years. Widowhood prevalence ranged from 14.1% in CHARLS to 53.9% in HAALSI and was consistently higher in women. Spousal loss was associated with poorer memory (multivariable-adjusted {beta} = -0.07, 95% CI -0.12 to -0.01) and executive function (multivariable-adjusted {beta} = -0.08, 95% CI -0.13 to -0.03) in the meta-analysis, with no significant associations for orientation or language. While results were generally consistent in five cohorts, the ELSA showed divergent patterns (orientation: {beta} = 0.10, 95% CI 0.06 to 0.13; memory: {beta} = 0.05, 95% CI 0.02 to 0.08; language: {beta} = 0.16, 95% CI 0.12 to 0.19). Sex-stratified analyses indicated poorer executive function among men (multivariable-adjusted {beta} = -0.14, 95% CI -0.19 to -0.08) and poorer memory among women (multivariable-adjusted {beta} = -0.07, 95% CI -0.14 to -0.01) following widowhood. Nonlinear age-related effects on cognition were observed in ELSA, LASI, and HAALSI. Higher education, internet use, and BMI were negatively associated with the risk of cognitive decline among widowed participants. Conclusions. Spousal loss is associated with domain- and sex-specific differences in cognitive performance, with substantial heterogeneity across study populations. Future research should integrate biopsychosocial markers to develop context-sensitive interventions for widowed older adults.

Objective: Cross-sectional studies indicate associations between self-reported social media use and adolescent wellbeing outcomes. We aimed to evaluate longitudinal associations of objectively measured smartphone and social media use with psychosocial wellbeing. Design: Observational study with one year of follow-up Setting: High schools in Finland from 2022 to 2023 Population: 259 adolescent girls (mean age 16.3 years at baseline) Main outcome measures: screenshots depicting smartphone and social media use, Bergen Social Media Addiction Scale (BSMAS), Generalized Anxiety Disorder-7 questionnaire, Body Appreciation Scale 2 (BAS-2) and visual analogue scales (VAS) of mood, tiredness, and loneliness Results: Across one year of follow-up, anxiety, body appreciation, and mood improved, but possible social media addiction increased from 15% to 17%. Social media addiction at baseline was associated with increased anxiety (r=0.29, p<0.001), lower body appreciation (r=-0.15, p=0.022), and more loneliness (r=0.20, p=0.001) at follow-up. Anxiety at baseline was associated with social media addiction at follow-up (r=0.19, p=0.005). The highest quartile of TikTok users reported more social media addiction (BSMAS 19 [IQR 16-21] vs. 17 [IQR 14-20]; p=0.009) and lower body appreciation (BAS-2 32 [IQR 28-38] vs. 35 [IQR 29-40]; p=0.003) than did others. The highest quartile of Snapchat users reported more social media addiction (BSMAS 19 [IQR 15-21] vs. 17 [IQR 14-20]; p=0.007) and tiredness (VAS 21 [IQR 13-32] vs. 26 [IQR 15-35]; p=0.049) than did others. Conclusions: Consistent with cross-sectional studies, social media addiction was associated with poorer psychosocial outcomes across follow-up. Policies to protect adolescents from social media addiction are urgently needed.

Background: Sleep debt and irregular sleep patterns are highly prevalent amongst adolescents. However, whether the absence of these sleep behaviours protects against subsequent depression remains unclear. Here, we examined the association of sleep debt, weekend catch-up sleep (WCS), and social jetlag (SJL) in adolescence with depression in young adulthood and identified underlying biopsychosocial mechanisms. Methods: Secondary data analyses were conducted using the Avon Longitudinal Study of Parents and Children. Bedtimes and wake-up times on school days and weekends (i.e., sleep duration) and sleep need were self-reported at 15 years. This was used to generate sleep debt (sleep need minus school day sleep duration), WCS (weekend sleep duration minus school day sleep duration), and SJL (absolute difference in the midpoint of sleep times between school days and weekends). Depression was assessed at 24 years with the Clinical Interview Schedule-Revised. Common mental health symptoms, biological, and school-related factors at 17 years were the mediators. Results: Logistic regression analyses revealed that greater WCS (adjusted odds ratio [AOR]=0.90; 95% CI=0.84-0.97; p=0.004) and lower sleep debt (AOR=1.10; 95% confidence interval [CI]=1.03-1.18; p=0.005) at age 15 reduced the likelihood of depression at 24 years. Irritability at 17 years partially mediated the relationship between sleep debt and depression (bias-corrected estimate=0.003; 95% CI=0.002-0.004; p<0.001). Conclusions: Adolescents who experience less sleep debt (i.e., less discrepancies between their actual sleep and their perceived sleep need) and those who extend their sleep duration on weekends are at reduced risk for depression in young adulthood. These findings underscore the need for greater opportunities for adolescents to obtain more hours of sleep to protect them against later poor mental health outcomes, such as depression. Keywords: Sleep; longitudinal studies; depression; ALSPAC

Background Home Test-to-Treat (HTTT) programs deliver timely antiviral treatment for acute respiratory infections, including COVID-19 and influenza, through at-home testing and telehealth. Because access is often measured by visit occurrence, variation in how and when care is delivered may be overlooked. We hypothesized that telehealth access follows distinct process-based patterns. Methods We analyzed de-identified encounters from the national HTTT program (September 2023-July 2024); 6,213 of 8,160 eligible individuals remained after exclusions for missing data. Phenotypes were derived by k-means clustering of standardized variables capturing encounter timing, modality preference, process duration, and sociodemographic and digital access attributes. Ten-day surveys assessed symptom duration and healthcare utilization. Results Three phenotypes emerged: Delayed/Disrupted Access (n = 1,537; 24.7%), Digitally Engaged but Socioeconomically Vulnerable (n = 1,460; 23.5%), and Mainstream Access and Efficient Utilization (n = 3,216; 51.8%). Mean process duration differed (15.93 [SD 3.84] vs 3.69 [3.31] vs 2.87 [2.41] hours; p < 0.001). Synchronous preference was lowest in the Digitally Engaged group (22.9%); antiviral prescribing was high (88.6%-91.9%). Among 10-day respondents (n = 1,023), symptom duration did not differ. Emergency department visits were most frequent in the Digitally Engaged group (2.3% vs 0.0% and 0.5%; p = 0.02) and urgent care in the Delayed/Disrupted group (5.8% vs 4.1% vs 2.0%; p = 0.02). Conclusions Telehealth use in a national HTTT program formed distinct phenotypes defined by timing, modality, and care-process efficiency. Evaluating equity requires attention to how and when care is delivered, not simply whether it occurred.

Background: Preparing tumor board patient summaries is time intensive. Large-language-model based systems may automate summarization but require real-world evaluation prior to clinical use. We performed an exploratory retrospective evaluation of the Microsoft Healthcare Agent Orchestrator (HAO), deployed in a Mayo Clinic controlled staged environment, to generate tumor board-style patient summaries from retrospective Electronic Health Record (EHR) notes. Methods: HAO generated summaries for breast, hepatobiliary, and neuro-oncology tumor board cases using up to the most recent 1,000 clinical notes. Clinician reviewers evaluated outputs via REDCap surveys across perceived factuality, completeness, clarity/conciseness, temporal cohesion, comparative performance, safety, and clinical utility (0-4 Likert scale). Reviewers were permitted to query the HAO chat interface to address missing details. Automated factuality was assessed using TBFact (bidirectional entailment), reporting precision and recall against available reference summaries. Results: Among 57 survey responses from 5 different physicians, mean scores exceeded 2.8 across domains, with medians of 3 for most axes. In an exploratory comparison, oncology fellows required less time to review HAO-generated summaries than to manually generate patient summaries (mean difference 13.57 minutes per patient, p<0.001), although this difference may be influenced by prior familiarity with the same cases; 96% of survey responses indicated that HAO would save time. TBFact evaluations showed higher recall than precision across domains, consistent with broad capture of reference content alongside additional content that was not present in gold-standard summaries. Attribution was viewed favorably but showed issues with primary-source specificity and link reliability. Conclusions: In a controlled Mayo environment, HAO demonstrated moderate performance and was associated with reduced review time for tumor board preparation. These findings are promising but preliminary and do not establish clinical safety, noninferiority to manual review, or readiness for routine clinical use. Limitations, including verbosity, specialty-specific content gaps, and inconsistent attribution, highlight the need for iterative refinement and further evaluation.

Longitudinal evaluations of national electronic health record repositories often track document counts alone, obscuring changes in content size, structure and standards implementation. We decomposed growth in the Estonian Health Information System across document counts, per-document size, section-level structure and version uptake in a 10% random population sample of 4.97 million HL7 Clinical Document Architecture Release 2 documents from 147,819 patients, spanning 2012--2019 and four prespecified document types. Growth patterns differed by document type. Inpatient summaries increased 48.5% in total content volume despite a 2.4% decline in document counts. Section presence and within-section content were highly skewed; 44.6% of 892 data locations carried one fixed value. Code-system diversity increased from 45 to 79, and version uptake took years: inpatient summaries reached 80% organisational uptake after a median 44 months (95% CI 11--78). This decomposition can guide extraction pipelines, secondary use and standards governance in CDA- and FHIR-based repositories.

Background: Black and Latino adults in the United States experience a disproportionate burden of cardiometabolic conditions due to interacting behavioral, social, and structural drivers of health. Less is known about the impact of integrating digital health tools into CHW-led interventions to improve cardiometabolic health. This trial evaluates a multilevel community-digital health promotion model delivered by CHWs to improve service utilization, health behaviors and cardiometabolic health among Black and Latino adults. Methods: This community-partnered trial uses a randomized delayed-control group with a phased recruitment design. Four cohorts (N = 664) are enrolled through three community-based organizations (CBOs). Eligible participants are 18 years who self-identify as Black or Latino, and have prediabetes/diabetes, hypertension, or overweight/obesity. Participants are allocated to either (1) a multilevel intervention consisting of CBO and CHW capacity building combined with individualized CHW-led lifestyle coaching and group activities supported by digital tools, or (2) a delayed control group receiving SMS-only cardiometabolic health education. Data collected at baseline, 6, 9, and 18 months include surveys and health metrics. Qualitative data are collected from participants and community partners to assess intervention acceptability, implementation facilitators and barriers, and sustainability. Results: The primary outcome is health service utilization at 6 and 9 months. Secondary outcomes include health behaviors, health metrics, and social determinants of health. Sustainability of health behaviors and health metrics is assessed at 18 months. Conclusions: Findings will provide evidence to inform scalable, sustainable community-digital health models for CHW-supported cardiometabolic health interventions in underserved communities.

Objective: Hispanic/Latinx populations in the U.S. experience higher rates of chronic disease linked to physical inactivity, yet digital health interventions remain largely inaccessible to more than 16 million Hispanic/Latinx adults with limited English proficiency. While large language models (LLMs) offer scalable personalization, their use in non-English behavioral coaching is unexplored. This study introduces MHC-Coach-ES, a Spanish-language LLM fine-tuned on the Transtheoretical Model (TTM) of behavior change. Materials and Methods: We fine-tuned Llama 3-70B-Instruct using a two-stage pipeline. First, the model was adapted to Spanish health and motivational language using a 2.21-million-token corpus. Second, it was instruction-tuned on 3,268 translated human written messages to align the model with the Transtheoretical Model (TTM) of Behavioral Change. We compared MHC-Coach-ES with Llama 3-70B-Instruct and translated human-expert messages using a forced-choice preference survey (N = 77) and blinded expert review (N = 2). Results: Spanish-speaking participants significantly preferred MHC-Coach-ES messages over translated human-expert messages (81% preference, P<0.001). Linguistic analysis showed that MHC-Coach-ES produced more temporally anchored messages than the base model (65% vs. 20%), while maintaining readability. In blinded evaluation, clinical experts rated MHC-Coach-ES higher for alignment with Transtheoretical Model stages than human-expert messages (4.83 vs. 4.38 out of 5). The base model also outperformed translated expert messages across preference and expert ratings. Conclusions: Generative AI can operationalize behavioral science frameworks in Spanish, offering a scalable approach to reducing health disparities. The strong performance of both MHC-Coach-ES and the base model highlights the promise of generative and personalized approaches over translation-based localization for theory-driven behavioral interventions.