Assessment of Neurodevelopment in Infants With and Without Exposure to Asymptomatic or Mild Maternal SARS-CoV-2 Infection During Pregnancy

Importance

Associations between prenatal SARS-CoV-2 exposure and neurodevelopmental outcomes have substantial public health relevance. A previous study found no association between prenatal SARS-CoV-2 infection and parent-reported infant neurodevelopmental outcomes, but standardized observational assessments are needed to confirm this finding.

Objective

To assess whether mild or asymptomatic maternal SARS-CoV-2 infection vs no infection during pregnancy is associated with infant neurodevelopmental differences at ages 5 to 11 months.

Design, Setting, and Participants

This cohort study included infants of mothers from a single-site prospective cross-sectional study (COVID-19 Mother Baby Outcomes [COMBO] Initiative) of mother-infant dyads and a multisite prospective cohort study (Epidemiology of Severe Acute Respiratory Syndrome Coronavirus 2 in Pregnancy and Infancy [ESPI]) of pregnant individuals. A subset of ESPI participants was subsequently enrolled in the ESPI COMBO substudy. Participants in the ongoing COMBO study were enrolled beginning on May 26, 2020; participants in the ESPI study were enrolled from May 7 to November 3, 2021; and participants in the ESPI COMBO substudy were enrolled from August 2020 to March 2021. For the current analysis, infant neurodevelopment was assessed between March 2021 and June 2022. A total of 407 infants born to 403 mothers were enrolled (204 from Columbia University Irving Medical Center in New York, New York; 167 from the University of Utah in Salt Lake City; and 36 from the University of Alabama in Birmingham). Mothers of unexposed infants were approached for participation based on similar infant gestational age at birth, date of birth, sex, and mode of delivery to exposed infants.

Exposures

Maternal symptomatic or asymptomatic SARS-CoV-2 infection.

Main Outcomes and Measures

Infant neurodevelopment was assessed using the Developmental Assessment of Young Children, second edition (DAYC-2), adapted for telehealth assessment. The primary outcome was age-adjusted standard scores on 5 DAYC-2 subdomains: cognitive, gross motor, fine motor, expressive language, and receptive language.

Results

Among 403 mothers, the mean (SD) maternal age at delivery was 32.1 (5.4) years; most mothers were of White race (240 [59.6%]) and non-Hispanic ethnicity (253 [62.8%]). Among 407 infants, 367 (90.2%) were born full term and 212 (52.1%) were male. Overall, 258 infants (63.4%) had no documented prenatal exposure to SARS-CoV-2 infection, 112 (27.5%) had confirmed prenatal exposure, and 37 (9.1%) had exposure before pregnancy or at an indeterminate time. In adjusted models, maternal SARS-CoV-2 infection during pregnancy was not associated with differences in cognitive (β = 0.31; 95% CI, -2.97 to 3.58), gross motor (β = 0.82; 95% CI, -1.34 to 2.99), fine motor (β = 0.36; 95% CI, -0.74 to 1.47), expressive language (β = -1.00; 95% CI, -4.02 to 2.02), or receptive language (β = 0.45; 95% CI, -2.15 to 3.04) DAYC-2 subdomain scores. Trimester of exposure and maternal symptom status were not associated with DAYC-2 subdomain scores.

Conclusions and Relevance

In this study, results of a novel telehealth-adapted observational neurodevelopmental assessment extended a previous finding of no association between prenatal exposure to maternal SARS-CoV-2 infection and infant neurodevelopment. Given the widespread and continued high prevalence of COVID-19, these data offer information that may be helpful for pregnant individuals who experience asymptomatic or mild SARS-CoV-2 infections.

Psychiatry training in autism spectrum disorder and intellectual disability: Ongoing gaps and emerging opportunities

LAY ABSTRACT

Children, adolescents, and adults with autism spectrum disorder and intellectual disability experience high rates of co-occurring psychiatric conditions throughout their lifetime. However, there is a shortage of psychiatrists to treat these populations. We evaluated how much education psychiatrists-in-training receive on how to care for individuals with autism spectrum disorder/intellectual disability. We found that in many psychiatry programs, residents receive limited training experiences in autism spectrum disorder/intellectual disability involving lectures and patient contact and that psychiatry program directors would benefit from more resources to strengthen education in autism spectrum disorder/intellectual disability.

Association of Birth During the COVID-19 Pandemic With Neurodevelopmental Status at 6 Months in Infants With and Without In Utero Exposure to Maternal SARS-CoV-2 Infection

IMPORTANCE

Associations between in utero exposure to maternal SARS-CoV-2 infection and neurodevelopment are speculated, but currently unknown.

OBJECTIVE

To examine the associations between maternal SARS-CoV-2 infection during pregnancy, being born during the COVID-19 pandemic regardless of maternal SARS-CoV-2 status, and neurodevelopment at age 6 months.

DESIGN, SETTING, AND PARTICIPANTS

A cohort of infants exposed to maternal SARS-CoV-2 infection during pregnancy and unexposed controls was enrolled in the COVID-19 Mother Baby Outcomes Initiative at Columbia University Irving Medical Center in New York City. All women who delivered at Columbia University Irving Medical Center with a SARS-CoV-2 infection during pregnancy were approached. Women with unexposed infants were approached based on similar gestational age at birth, date of birth, sex, and mode of delivery. Neurodevelopment was assessed using the Ages & Stages Questionnaire, 3rd Edition (ASQ-3) at age 6 months. A historical cohort of infants born before the pandemic who had completed the 6-month ASQ-3 were included in secondary analyses.

EXPOSURES

Maternal SARS-CoV-2 infection during pregnancy and birth during the COVID-19 pandemic.

MAIN OUTCOMES AND MEASURES

Outcomes were scores on the 5 ASQ-3 subdomains, with the hypothesis that maternal SARS-CoV-2 infection during pregnancy would be associated with decrements in social and motor development at age 6 months.

RESULTS

Of 1706 women approached, 596 enrolled; 385 women were invited to a 6-month assessment, of whom 272 (70.6%) completed the ASQ-3. Data were available for 255 infants enrolled in the COVID-19 Mother Baby Outcomes Initiative (114 in utero exposed, 141 unexposed to SARS-CoV-2; median maternal age at delivery, 32.0 [IQR, 19.0-45.0] years). Data were also available from a historical cohort of 62 infants born before the pandemic. In utero exposure to maternal SARS-CoV-2 infection was not associated with significant differences on any ASQ-3 subdomain, regardless of infection timing or severity. However, compared with the historical cohort, infants born during the pandemic had significantly lower scores on gross motor (mean difference, -5.63; 95% CI, -8.75 to -2.51; F1,267 = 12.63; P<.005), fine motor (mean difference, -6.61; 95% CI, -10.00 to -3.21; F1,267 = 14.71; P < .005), and personal-social (mean difference, -3.71; 95% CI, -6.61 to -0.82; F1,267 = 6.37; P<.05) subdomains in fully adjusted models.

CONCLUSIONS AND RELEVANCE

In this study, birth during the pandemic, but not in utero exposure to maternal SARS-CoV-2 infection, was associated with differences in neurodevelopment at age 6 months. These early findings support the need for long-term monitoring of children born during the COVID-19 pandemic.

Interpretable Neuron Structuring with Graph Spectral Regularization

While neural networks are powerful approximators used to classify or embed data into lower dimensional spaces, they are often regarded as black boxes with uninterpretable features. Here we propose Graph Spectral Regularization for making hidden layers more interpretable without significantly impacting performance on the primary task. Taking inspiration from spatial organization and localization of neuron activations in biological networks, we use a graph Laplacian penalty to structure the activations within a layer. This penalty encourages activations to be smooth either on a predetermined graph or on a feature-space graph learned from the data via co-activations of a hidden layer of the neural network. We show numerous uses for this additional structure including cluster indication and visualization in biological and image data sets.

Abstract 5306: SAUCIE: Sparse autoencoder for unsupervised clustering, imputation, and embedding

Handling the vast amounts of single-cell RNA-sequencing and CyTOF data, now being generated in patient cohorts, presents a computational challenge due to the noise, complexity, sparsity and batch effects present. Here, we propose a unified deep neural network approach to automatically process and extract structure from these massive datasets. Our unsupervised architecture, called SAUCIE (Sparse Autoencoder for Unsupervised Clustering, Imputation, and Embedding), simultaneously performs several key tasks for single-cell data analysis including 1) clustering, 2) batch correction, 3) visualization, and 4) denoising/imputation. SAUCIE is trained to recreate its own input after reducing its dimensionality in a 2-D embedding layer which can be used to visualize the data. Additionally, it uses two novel regularizations: (1) an information dimension regularization to penalize entropy as computed on normalized activation values of the layer, and thereby encourage binary-like encodings that are amenable to clustering and (2) a Maximal Mean Discrepancy penalty to correct batch effects. Thus SAUCIE has a single architecture that denoises, batch-corrects, visualizes and clusters data using a unified representation. We show results on artificial data where ground truth is known, as well as mass cytometry data from patients undergoing immunotherapy, dengue-infected patients, and single-cell RNA-sequencing data from embryonic mouse brain.

Citation Format: Matthew Amodio, Krishnan Srinivasan, David van Dijk, Hussein Mohsen, Kristina Yim, Rebecca Muhle, Kevin Moon, Ruth Montgomery, James Noonan, Guy Wolf, Smita Krishnaswamy. SAUCIE: Sparse autoencoder for unsupervised clustering, imputation, and embedding [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 5306.

Modular multipin electrodes for comfortable dry EEG

Electrode and cap concepts for continuous and ubiquitous monitoring of brain activity will open up new fields of application and contribute to increased use of electroencephalography (EEG) in clinical routine, neurosciences, brain-computer-interfacing and out-of-the-lab monitoring. However, mobile and unobtrusive applications are currently hindered by the lack of applicable convenient and reliable electrode and cap systems. We propose a novel modular electrode concept based on a flexible polymer substrate, coated with electrically conductive metallic films. The overall concept enables design adaptation to different head regions and cap designs. We describe the single modules of the system and investigate the influence of electrode pin number, coating material and adduction force on electrode-skin impedance and perceived wearing comfort. Our results contribute to rapid and comfortable multichannel dry EEG.

The genetics of autism

Autism is a complex, behaviorally defined, static disorder of the immature brain that is of great concern to the practicing pediatrician because of an astonishing 556% reported increase in pediatric prevalence between 1991 and 1997, to a prevalence higher than that of spina bifida, cancer, or Down syndrome. This jump is probably attributable to heightened awareness and changing diagnostic criteria rather than to new environmental influences. Autism is not a disease but a syndrome with multiple nongenetic and genetic causes. By autism (the autistic spectrum disorders [ASDs]), we mean the wide spectrum of developmental disorders characterized by impairments in 3 behavioral domains: 1) social interaction; 2) language, communication, and imaginative play; and 3) range of interests and activities. Autism corresponds in this article to pervasive developmental disorder (PDD) of the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition and International Classification of Diseases, Tenth Revision. Except for Rett syndrome--attributable in most affected individuals to mutations of the methyl-CpG-binding protein 2 (MeCP2) gene--the other PDD subtypes (autistic disorder, Asperger disorder, disintegrative disorder, and PDD Not Otherwise Specified [PDD-NOS]) are not linked to any particular genetic or nongenetic cause. Review of 2 major textbooks on autism and of papers published between 1961 and 2003 yields convincing evidence for multiple interacting genetic factors as the main causative determinants of autism. Epidemiologic studies indicate that environmental factors such as toxic exposures, teratogens, perinatal insults, and prenatal infections such as rubella and cytomegalovirus account for few cases. These studies fail to confirm that immunizations with the measles-mumps-rubella vaccine are responsible for the surge in autism. Epilepsy, the medical condition most highly associated with autism, has equally complex genetic/nongenetic (but mostly unknown) causes. Autism is frequent in tuberous sclerosis complex and fragile X syndrome, but these 2 disorders account for but a small minority of cases. Currently, diagnosable medical conditions, cytogenetic abnormalities, and single-gene defects (eg, tuberous sclerosis complex, fragile X syndrome, and other rare diseases) together account for <10% of cases. There is convincing evidence that "idiopathic" autism is a heritable disorder. Epidemiologic studies report an ASD prevalence of approximately 3 to 6/1000, with a male to female ratio of 3:1. This skewed ratio remains unexplained: despite the contribution of a few well characterized X-linked disorders, male-to-male transmission in a number of families rules out X-linkage as the prevailing mode of inheritance. The recurrence rate in siblings of affected children is approximately 2% to 8%, much higher than the prevalence rate in the general population but much lower than in single-gene diseases. Twin studies reported 60% concordance for classic autism in monozygotic (MZ) twins versus 0 in dizygotic (DZ) twins, the higher MZ concordance attesting to genetic inheritance as the predominant causative agent. Reevaluation for a broader autistic phenotype that included communication and social disorders increased concordance remarkably from 60% to 92% in MZ twins and from 0% to 10% in DZ pairs. This suggests that interactions between multiple genes cause "idiopathic" autism but that epigenetic factors and exposure to environmental modifiers may contribute to variable expression of autism-related traits. The identity and number of genes involved remain unknown. The wide phenotypic variability of the ASDs likely reflects the interaction of multiple genes within an individual's genome and the existence of distinct genes and gene combinations among those affected. There are 3 main approaches to identifying genetic loci, chromosomal regions likely to contain relevant genes: 1) whole genome screens, searching for linkage of autism to shared genetic markers in populations of multiplex families (families with >1 affected family member; 2) cytogenetic studies that may guide molecular studies by pointing to relevant inherited or de novo chromosomal abnormalities in affected individuals and their families; and 3) evaluation of candidate genes known to affect brain development in these significantly linked regions or, alternatively, linkage of candidate genes selected a priori because of their presumptive contribution to the pathogenesis of autism. Data from whole-genome screens in multiplex families suggest interactions of at least 10 genes in the causation of autism. Thus far, a putative speech and language region at 7q31-q33 seems most strongly linked to autism, with linkages to multiple other loci under investigation. Cytogenetic abnormalities at the 15q11-q13 locus are fairly frequent in people with autism, and a "chromosome 15 phenotype" was described in individuals with chromosome 15 duplications. Among other candidate genes are the FOXP2, RAY1/ST7, IMMP2L, and RELN genes at 7q22-q33 and the GABA(A) receptor subunit and UBE3A genes on chromosome 15q11-q13. Variant alleles of the serotonin transporter gene (5-HTT) on 17q11-q12 are more frequent in individuals with autism than in nonautistic populations. In addition, animal models and linkage data from genome screens implicate the oxytocin receptor at 3p25-p26. Most pediatricians will have 1 or more children with this disorder in their practices. They must diagnose ASD expeditiously because early intervention increases its effectiveness. Children with dysmorphic features, congenital anomalies, mental retardation, or family members with developmental disorders are those most likely to benefit from extensive medical testing and genetic consultation. The yield of testing is much less in high-functioning children with a normal appearance and IQ and moderate social and language impairments. Genetic counseling justifies testing, but until autism genes are identified and their functions are understood, prenatal diagnosis will exist only for the rare cases ascribable to single-gene defects or overt chromosomal abnormalities. Parents who wish to have more children must be told of their increased statistical risk. It is crucial for pediatricians to try to involve families with multiple affected members in formal research projects, as family studies are key to unraveling the causes and pathogenesis of autism. Parents need to understand that they and their affected children are the only available sources for identifying and studying the elusive genes responsible for autism. Future clinically useful insights and potential medications depend on identifying these genes and elucidating the influences of their products on brain development and physiology.

Intrauterine bacterial inoculation induces labor in the mouse by mechanisms other than progesterone withdrawal

We tested the hypothesis that progesterone (P(4)) withdrawal is the primary mechanism by which intrauterine bacteria induce preterm labor in mice. CD-1 mice on Day 14.5 of a 19- to 20-day gestation were subjected to one of four treatments: 1) intrauterine injection of sterile medium, 2) intrauterine injection of 10(6) heat-killed Escherichia coli bacteria, 3) intrauterine injection of 10(9) heat-killed E. coli, or 4) ovariectomy. Mice were then killed at four time points from 0.75 to 11 h after surgery for serum collection. Separately, animals were pretreated either with s.c. P(4) or with vehicle 2 h before ovariectomy or high-dose bacterial inoculation. Ovariectomy led to a rapid fall in serum P(4) levels of 60% by 1 h and 81% by 8 h compared with levels in controls (P < 0.001). In contrast, intrauterine inoculation with 10(9) bacteria led to a more modest decline in P(4) of only 28% by 8 h (P = 0.24, which was no different from that of 10(6) bacteria, an inoculum below the threshold for preterm delivery). Despite significantly lower levels of P(4) in the ovariectomy group, time to delivery was significantly shorter with 10(9) bacteria intrauterine (24 +/- 5.6 h vs. 19 +/- 3.6 h, P = 0.03). Pretreatment with 1.5 mg P(4) per mouse prolonged the interval to delivery following both ovariectomy and high-dose bacteria, in association with pharmacologically elevated serum P(4) levels. In contrast, physiologic P(4) supplementation (0.375 mg/mouse) prolonged gestation only in the ovariectomy group. We conclude that withdrawal of endogenous P(4) is not the primary cause of labor following intrauterine bacterial inoculation in mice.

Identification of mammalian Sds3 as an integral component of the Sin3/histone deacetylase corepressor complex

Silencing of gene transcription involves local chromatin modification achieved through the local recruitment of large multiprotein complexes containing histone deacetylase (HDAC) activity. The mammalian corepressors mSin3A and mSin3B have been shown to play a key role in this process by tethering HDACs 1 and 2 to promoter-bound transcription factors. Similar mechanisms appear to be operative in yeast, in which epistasis experiments have established that the mSin3 and HDAC orthologs (SIN3 and RPD3), along with a novel protein, SDS3, function in the same repressor pathway. Here, we report the identification of a component of the mSin3-HDAC complex that bears homology to yeast SDS3, physically associates with mSin3 proteins in vivo, represses transcription in a manner that is partially dependent on HDAC activity, and enables HDAC1 catalytic activity in vivo. That key physical and functional properties are also shared by yeast SDS3 underscores the central role of the Sin3-HDAC-Sds3 complex in eukaryotic cell biology, and the discovery of mSds3 in mammalian cells provides a new avenue for modulating the activity of this complex in human disease.

Role for N-CoR and histone deacetylase in Sin3-mediated transcriptional repression

Normal mammalian growth and development are highly dependent on the regulation of the expression and activity of the Myc family of transcription factors. Mxi1-mediated inhibition of Myc activities requires interaction with mammalian Sin3A or Sin3B proteins, which have been purported to act as scaffolds for additional co-repressor factors. The identification of two such Sin3-associated factors, the nuclear receptor co-repressor (N-CoR) and histone deacetylase (HD1), provides a basis for Mxi1/Sin3-induced transcriptional repression and tumour suppression.