Wesley K. Thompson’s research while affiliated with Laureate Institute for Brain Research and other places

Infections in utero and early childhood are associated with an increased epilepsy risk; however, confounding by familial predisposition has not been adequately accounted for in previous studies. We aimed to assess the epilepsy risk attributable to infections in utero and early childhood by performing population-based and sibling-comparison analyses to account for residual and unmeasured familial confounding factors. This nationwide birth cohort study included 2,609,289 individuals born 2001–2016 in Taiwan. Maternal infection during pregnancy and early childhood infection during the first year of life were defined. Maternal pre-pregnancy infection was used as negative control. In the population analyses, offspring exposed to any maternal infection during pregnancy had an increased epilepsy risk (hazard ratio (HR) = 1.36, 95% confidence interval (CI):1.27–1.45). However, the association with maternal infection was attenuated to the null (HR = 1.11, 95% CI:0.98–1.27), except for maternal infection in sepsis (HR = 2.54, 95% CI:1.74–3.70) and central nervous system (HR = 24.59, 95% CI:3.28–184.23), in the sibling analyses. The association of maternal pre-pregnancy infection with offspring epilepsy was observed in the population analyses, but not in the sibling analyses. Individuals exposed to childhood infection had an increased epilepsy risk (HR = 1.49, 95% CI:1.45–1.54) in the population analyses; the association was still observed in the sibling analyses (HR = 1.31, 95% CI:1.23–1.40). The association between maternal infection during pregnancy and epilepsy risk in the offspring appears largely because of familial confounding factors. Infections during early childhood may play a causal role in the subsequent epilepsy risk.

Objectives Neurodevelopmental disorders have significant public health impacts, and novel approaches to understanding these disorders are greatly needed. Social connectedness, including relationships with parents and peers as well as family and school environments, may serve as a protective factor for neurodivergent youth. Neural networks that support social processing could also influence outcomes for these individuals. Methods The current study used data from the large Adolescent Brain Cognitive DevelopmentSM (ABCD) Study (N = 11,878) to explore associations between social connectedness, brain functional connectivity, and subjective and objective measures of neurodevelopmental functioning in youth. Linear mixed effects models assessed links between social connectedness and neurodevelopmental functioning. Mediation models evaluated whether connectivity of the salience network mediated these associations. Results Authoritative parenting practices and involvement in extracurricular activities were linked to better neurodevelopmental functioning (ps ≤ 0.002), while family conflict was associated with worse functioning (ps ≤ 0.003). Salience network connectivity showed no significant associations with either social connectedness or neurodevelopmental functioning (ps > 0.01). Conclusions Although the sizes of the observed effects were small, our findings imply that fostering positive family relationships and encouraging extracurricular involvement may relate to improved outcomes for neurodivergent youth. Future work is needed to identify additional factors that influence neurodevelopmental functioning over time and to determine the neural mechanisms underlying these relationships.

Neuroticism is characterized by emotional instability and increased susceptibility to stress-related disorders. While traditionally treated as a unitary construct, growing evidence suggests that neuroticism is heterogeneous in both its genetic basis and its effects on social and health outcomes. To quantify this heterogeneity at the neurofunctional level, we analyzed resting-state functional connectivity (RSFC) in a large sample (n = 33,180) from the UK Biobank dataset. Using machine learning regression analysis, we identified RSFC patterns associated with item-level responses from the neuroticism scale of the Eysenck Personality Questionnaire. The pattern of RSFC associations across questionnaire items reflected genetically defined clusters (Worry and Depressed Affect), showing a significant correlation (r = 0.767, p < 0.001). It also aligned with psycholometrically derived factors (Anxiety/Tension and Worry/Vulnerability), reflecting a factor structure consistent with prior psychological studies. These findings were replicated in a separate MRI scan session from the UK biobank dataset. Associated connectivity was primarily observed in cognitive control, sensory integration, and self-referential processing networks. These neurofunctional signatures position RSFC as a robust intermediate phenotype, bridging genetic predisposition for neuroticism and psychological states. This framework could enhance precision in predicting psychiatric vulnerability and inform tailored therapeutic interventions.

Linear mixed-effects (LME) models are commonly used for analyzing longitudinal data. However, most applications of LME models rely on random intercepts or simple, e.g., stationary, covariance. Here, we extend the Fast and Efficient Mixed-Effects Algorithm (FEMA) and present FEMA-Long, a computationally tractable approach to flexibly modeling longitudinal covariance suitable for high-dimensional data. FEMA-Long can i) model unstructured covariance, ii) model non-linear fixed effects using splines, iii) discover time-dependent fixed effects with spline interactions, and iv) perform genome-wide association studies (GWAS) supporting discovery of time-dependent genetic effects. We applied FEMA-Long to perform a longitudinal GWAS with non-linear SNP-by-time interaction on length, weight, and body mass index of 68,273 infants with up to six measurements in the first year of life. We found dynamic patterns of random effects including time-varying heritability and correlations, as well as several genetic variants showing time-dependent effects, highlighting the applicability of FEMA-Long to enable novel discoveries.

Background: Adolescent mental health issues were surging during the COVID-19 pandemic. Yet it is unclear whether the pandemic amplified pre-existing vulnerabilities for psychiatric disorders. Methods: Using longitudinal data from the Adolescent Brain Cognitive Development (ABCD) Study (n = 7,560, 2-3 waves of assessments before the pandemic and 2-3 waves after the first nation-wide pandemic lockdown), we evaluated associations of the pandemic, genetic liabilities to psychiatric disorders, and their interactions with 20 different measures of psychopathology. Genomic common factor models aggregated genomic effects across eight psychiatric disorders, summarized into four latent factors. Analyses were stratified by genetic ancestry and sex. Results: In adolescents of European-like ancestry, each 1 standard deviation increase in Neurodevelopmental (ND) or Internalizing (INT) PRS was significantly associated with an increase in most psychopathologies by 3% to 19%. After controlling for individual PRS, pandemic periods were significantly associated with accelerated rates in parent-reported Child Behavior Checklist (CBCL) withdrawn depressed and rule-breaking syndrome scores, CBCL DSM-oriented conduct, somatic, and attention deficit/hyperactivity problems, and all corresponding youth-reported Brief Problem Monitor (BPM-Y) scores. In sex-stratified analysis, CBCL DSM-oriented affective problem scores significantly worsened in the early pandemic among females (21% increase; 95% CI 13%-29%; P = 2.6e-8) but not males. Females showed stronger associations between INT PRS and the rate of increase in CBCL DSM-oriented affective problem scores (15%; 95% CI 10%-21%; P = 6.4e-10), compared to males (10%; 95% CI 6%-15%; P = 4.2e-6). The multiplicative interactions between PRS and pandemic periods were at most trending, showing positive interactions between ND PRS and the early pandemic among females for CBCL conduct (15%; 95% CI 7%-23%; P = 5.17e-5) and aggressive behavior scores (9%; 95% CI 4%-14%; P = 5.09e-4). Conclusion: A wide range of adolescent mental health symptoms intensified during the pandemic period. Both genetic vulnerabilities and pandemic-related factors were associated with increased psychiatric symptoms. The genetic liability and the pandemic periods were associated with mental health issues independently, meaning genetically at-risk individuals saw a higher relative increase in mental health problems during the pandemic. Females exhibited higher levels of mental health symptoms and more sustained increases across the duration of the pandemic compared to males. Youth with genetic vulnerability to neurodevelopmental phenotypes require special attention due to heightened mental health risks during stressors like COVID-19.

Recent advances in large language models (LLMs) have prompted a frenzy in utilizing them as universal translators for biomedical terms. However, the black box nature of LLMs has forced researchers to rely on artificially designed benchmarks without understanding what exactly LLMs encode. We demonstrate that pretrained LLMs can already explain up to 51% of the genetic correlation between items from a psychometrically-validated neuroticism questionnaire, without any fine-tuning. For psychiatric diagnoses, we found disorder names aligned better with genetic relationships than diagnostic descriptions. Our results indicate the pretrained LLMs have encodings mirroring genetic architectures. These findings highlight LLMs potential for validating phenotypes, refining taxonomies, and integrating textual and genetic data in mental health research.

Importance Treatment-resistant depression (TRD) is a major challenge in mental health, affecting a significant number of patients and leading to considerable burdens. The etiological factors contributing to TRD are complex and not fully understood. Objective To investigate the genetic factors associated with TRD using polygenic scores (PGS) across various traits and explore their potential role in the etiology of TRD using large-scale genomic data from the All of Us (AoU) Research Program. Design, Setting, and Participants This study was a cohort design with observational data from participants in the AoU Research Program who have both electronic health records and genomic data. Data analysis was performed from March 27 to October 24, 2024. Exposures PGS for 61 unique traits from 7 domains. Main Outcomes and Measures Logistic regressions to test if PGS was associated with treatment-resistant depression (TRD) compared with treatment-responsive major depressive disorder (trMDD). Cox proportional hazard model was used to determine if the progressions from MDD to TRD were associated with PGS. Results A total of 292 663 participants (median [IQR] age, 57 (41-69) years; 175 981 female [60.1%]) from the AoU Research Program were included in this analysis. In the discovery set (124 945 participants), 11 of the selected PGS were found to have stronger associations with TRD than with trMDD, encompassing PGS from domains in education, cognition, personality, sleep, and temperament. Genetic predisposition for insomnia (odds ratio [OR], 1.11; 95% CI, 1.07-1.15) and specific neuroticism (OR, 1.11; 95% CI, 1.07-1.16) traits were associated with increased TRD risk, whereas higher education (OR, 0.88; 95% CI, 0.85-0.91) and intelligence (OR, 0.91; 95% CI, 0.88-0.94) scores were protective. The associations held across different TRD definitions (meta-analytic R ² >83%) and were consistent across 2 other independent sets within AoU (the whole-genome sequencing Diversity dataset, 104 388, and Microarray dataset, 63 330). Among 28 964 individuals followed up over time, 3854 developed TRD within a mean of 944 days (95% CI, 883-992 days). All 11 previously identified and replicated PGS were found to be modulating the conversion rate from MDD to TRD. Conclusions and Relevance Results of this cohort study suggest that genetic predisposition related to neuroticism, cognitive function, and sleep patterns had a significant association with the development of TRD. These findings underscore the importance of considering psychosocial factors in managing and treating TRD. Future research should focus on integrating genetic data with clinical outcomes to enhance understanding of pathways leading to treatment resistance.