This issue of the Journal brings together papers that address issues relevant to understanding risk factors, biomarkers, and predictors of treatment response for psychiatric illnesses that commonly emerge early in life, such as autism spectrum disorder (ASD), attention deficit hyperactivity disorder (ADHD), and obsessive-compulsive disorder (OCD). DSM-5-TR explicitly categorizes ASD and ADHD as neurodevelopmental disorders whereas OCD, which also commonly presents during childhood and adolescence, is not “officially” considered a neurodevelopmental disorder. We begin the issue with a review by Dr. Mara Parellada and members of the APA Biomarkers Taskforce (1) that evaluates the evidence for the reliable use of ASD biomarkers to assess treatment outcomes in ASD clinical trials. Following the theme of ASD biomarkers, one of the research articles in this issue presents data that more specifically assess the validity of EEG biomarkers for use in ASD research. Another study in this issue examines brain structure in relation to ASD, reporting on sex-related neuroanatomy and gene expression patterns in males and females.
Also relevant to ASD and other disorders, we include an article that characterizes the prevalence of rare neurodevelopment disorder-related genetic variants in a large health care population and their association with ASD, ADHD, schizophrenia, and bipolar disorder diagnoses. Focusing on ADHD, another study characterizes the extent to which the polygenic risk for ADHD is also associated with other factors that have been linked to the development of ADHD. And the final article in this issue presents data from a clinical trial in OCD patients assessing pretreatment measures of resting state functional connectivity in relation to responses to psychotherapy.
Assessing the Validity of EEG Biomarkers for Use in ASD Research
Biomarker development is critical for psychiatric research efforts that are focused on developing strategies aimed at predicting outcomes related to prognosis and course of illness, as well as in guiding treatment selection and predicting treatment responses. It is the hope that reliable biomarkers will be sufficiently robust and reproducible so that the evidence base will eventually support their clinical use as part of personalized treatment approaches. Despite the enthusiasm for biomarkers, their use across different psychiatric disorders often yields inconsistent results. In this regard, Webb et al. (2) focus on attempting to validate the use of various EEG paradigms as reliable biomarkers for ASD research. The purpose of this study is to examine the psychometric properties of four EEG paradigms (i.e., resting-state EEG activity, face perception, motion perception, and visual evoked potentials) that are recommended for use in research by the Autism Biomarkers Consortium for Clinical Trials. Using EEG data collected at five sites from a sample of 280 children with ASD and 119 typically developing children, the researchers evaluated acquisition rates, construct performance, and within-subject stability for the four EEG paradigms. In general, the acquisition of the EEG data was good, as successful data were collected for greater than 70% of the ASD participants for the resting-state, face perception, and visual evoked potential paradigms; this was not the case for the motion perception paradigm. Similarly, the resting-state, face, and visual evoked paradigms demonstrated the best construct performance, i.e., the extent to which the EEG responses to the paradigms—and the differences in response between the ASD and control participants—were similar to those expected from prior work. Additionally, the EEG responses within individuals demonstrated moderate stability when assessed twice over a 6-week period and individual differences in EEG responses also correlated with some of the ASD-related phenotypic measures. From these data, the authors surmise that among these recommended paradigms, the evidence best supports efforts at further developing the EEG face response paradigm as a potential reliable ASD biomarker. In their editorial, Matthew Lerner and Talena Day from Stony Brook University (3) underscore the rigorous methods used in this study to assess the reliability of the various EEG paradigms and discuss the use of these and other biomarkers as potential outcome measures in ASD clinical trials.
Sex-Related Neuroanatomy Is Associated With ASD: Cognitive and Gene Expression Correlates
ASD is the prototypical neuropsychiatric neurodevelopmental disorder and in 2021 the prevalence of ASD in children was estimated to be 0.2%, with a fourfold increased prevalence in boys compared with girls (4). The mechanisms underlying the greater risk for boys are not known; however, evidence from various studies supports an association between male-related brain structure and ASD in both males and females. Floris and colleagues (5) use large structural imaging databases acquired from various sources that included the UK Biobank, the Autism Brain Imaging Data Exchange, and the Longitudinal European Autism Project in an attempt to deepen the understanding of the relation between male-related brain structure and ASD. The authors conceptualize ASD-associated brain alterations as an intermediate phenotype juxtaposed between circuit-specific gene expression patterns and behavior. Therefore, in ASD individuals, this research explores the relations among sex-related brain structure, patterns of regional brain gene expression, and measures of cognition. Machine learning techniques were employed to analyze structural imaging data demonstrating that the neurotypical features predictive of male brain structure were more predictive of males with ASD while brain structure associated with neurotypical females was found to be less predictive of ASD females. In females with ASD, greater male-related brain features were associated with increased deficits related to social sensitivity and face processing. By using data from the Allen Institute Human Brain Gene Expression Atlas, genes highly expressed in male-associated brain regions were selected for further analysis. Next, overlaps between these selected genes with known ASD-related upregulated genes were determined. When characterizing the genes in ASD females that were overexpressed in these regions, the investigators found an enrichment in genes related to prenatal excitatory cells and genes that are down regulated by estrogen. In contrast, in males with ASD, male-related regions were associated with genes related to prenatal microglia, radial glial cells, and genes that are upregulated by estrogen and dihydrotestosterone, a product of testosterone. These data begin to shed light on the molecular underpinnings of ASD and mechanisms that may underlie differences between males and females with ASD. Overall, these data support the theory that processes underlying the early sexual differentiation of the brain are involved in the etiology of autism. In her editorial, Dr. Leanna Hernandez (6) from UCLA provides a discussion of the findings and the complex methods used in the analyses. She also discusses the findings in relation to current theories aimed at explaining phenotypic differences in the presentation of ASD and the difference in ASD prevalence in males compared to females.
Rare Genetic Variants and Psychiatric-Related Neurodevelopment Disorders in a Large Health Care Population Sample
While in general, the genetics of psychiatric disorders are complex and polygenic in nature, rare single gene mutations have been demonstrated to underlie illness in a small proportion of psychiatric patients. Understanding the functional significance of these mutations is important because it may provide insights into molecular pathways that are more universally involved in mediating the pathophysiology of a specific disorder. Many psychiatric disorders, including ASD, ADHD, and schizophrenia, have their origins early in life in part due to the involvement of altered neurodevelopmental processes, and Shimelis et al. (7) focus on understanding the prevalence of rare pathogenic gene variants relevant to psychiatric-related neurodevelopmental disorders. Building on earlier work (8, 9), using a large sample from the Geisinger health care population, this study used exome sequencing data and electronic medical records from a cohort of 90,595 participants to determine the prevalence of 94 loss of function genetic variants known to be associated with neurodevelopmental disorders. As a next step, the penetrance of these pathogenic genetic variants, the extent to which they were associated with psychiatric illnesses, was determined. The findings revealed that 233/90,595 participants or 0.34% of the sample had pathogenic gene variants that involved 61 of the 94 preselected neurodevelopment-related genes. When assessing penetrance, the investigators found that a significantly higher percentage of individuals had a neurodevelopmental disorder diagnosis if they had at least one of the pathogenic variants, compared with those that did not have one of the pathogenic gene variants (34.3% vs. 14.6%). Of interest, a high degree of penetrance of at least 40% was found for eight genes. When the researchers also included anxiety, depression, and congenital anomalies into their penetrance analyses, they found that 71% of individuals with a pathogenic variant had a neurodevelopmental disorder, anxiety, depression, or a congenital anomaly. However, this was not especially discriminating as 60% of individuals without a genetic variant also had one of these disorders. Among the neurodevelopmental problems, intellectual disability was most associated with pathogenic variants (odds ratio=5.89), as were other neurological disorders and learning and communication disorders. A 4.56 odds ratio was found for ASD, however this finding did not survive multiple comparison correction. Regarding other psychiatric illnesses, having one of the genetic variants significantly increased the odds of having the diagnosis of schizophrenia/other psychotic disorders, bipolar disorder, or ADHD by two-to-threefold. In their editorial (10), Deepro Banerjee and Santhosh Girirajan from Penn State University discuss these findings, emphasizing the importance of studying the effects of deleterious variants in large, non-preselected populations. They point out that in contrast to samples selected for pathology, studies using unselected samples more accurately reflect the penetrance of genetic variants in the absence of selection biases. They also discuss the potential utility of population-based data like these in informing future clinical strategies such as a “genotype-first approach for diagnosis.”
How Does the Polygenic Risk for ADHD Interact With Other ADHD Risk Factors?
In addition to the heritable component of ADHD, which is estimated to be around 75%, numerous other birth-related, somatic, psychosocial, and demographic factors have been associated with increasing the risk to develop ADHD. Examples of these factors include low birth weight, maternal hypertension, infections, autoimmune disease, low income, and parental psychiatric disorders. It is interesting that some evidence suggests that the same genes that underlie the development of ADHD may also be related to the likelihood of being exposed to some of these risk factors. A more in-depth understanding of the degree to which these other risk factors are influenced by, or are independent from, ADHD polygenic risk genes is valuable for teasing apart the mechanisms underlying ADHD risk. In this regard, Brikell and colleagues (11) use data from a large Danish sample selected from the Psychiatric Research consortium (iPSCYH2012) to explore potential interactions between polygenetic gene scores and other ADHD risk factors. The sample studied consisted of 21,578 randomly selected individuals along with 13,697 individuals diagnosed with ADHD. In addition to polygenic risk scores, 24 other risk factors that were established from previous studies were examined. As expected, polygenic risk scores for ADHD predicted the likelihood of having ADHD such that individuals with the greatest genetic loading, in the highest polygenic risk score decile, had a 4.42 increased relative risk compared to individuals in the lowest polygenic risk score decile. In addition, although with very small odds ratios, the polygenic risks scores for ADHD were significantly associated with 12 of the 24 risk factors. These included such factors as being small for gestational age, maternal autoimmune disorder, infections, mild traumatic brain injury, and many of the family psychosocial factors. In contrast, birthweight, low APGAR scores, maternal hypertension, infection during pregnancy, and other illnesses were not associated with ADHD polygenic risk scores. When assessed in the current sample, 19 of the risk factors were found to be significantly associated with ADHD. The most prominent association for the birth-related category was low birth weight (relative risk=1.85); for the somatic category the most robust association was with epilepsy (relative risk=2.38); and in the psychosocial domain the strongest association was with low parental education level (relative risk=approximately 3.5). Of importance is that these associations between the risk factors and ADHD were not significantly weakened when performing analyses that covaried for ADHD polygenic risk scores or parental psychiatric history. When performing interaction analyses, only the interaction between polygenic risk scores and a history of maternal autoimmune disorders was significant (false discovery rate corrected) demonstrating that having a maternal history of autoimmune disorders modestly reduced the impact of the polygenic risk score on the likelihood of having ADHD. Taken together, the findings from this study demonstrate that the polygenic risk for developing ADHD is also associated with many of the risk factors associated with ADHD. The data suggest, however, that these risk factors may influence the likelihood of developing ADHD in a manner that is independent of the influence of the polygenic risk on developing ADHD. In her editorial, Dr. Sarah Kittel-Schneider from Universitat Wurzburg (12) discusses the findings from Brikell et al., emphasizing the complexity of the interactions between a multitude of genes and other developmental and environmental factors in conferring the risk to develop ADHD.
Resting-State Connectivity Predictors of Treatment Response to Psychotherapy in Obsessive-Compulsive Disorder Patients
Using resting-state fMRI connectivity measures, Russman-Block et al. (13) attempt to identify patterns of resting-state connectivity in patients with OCD that predict psychotherapy treatment responses. In a randomized clinical trial, the investigators compared the efficacy of 12 weeks of exposure response prevention (N=58) to an active comparator, stress management therapy (N=58), in adolescents and adults with OCD. Resting-state fMRI data was collected before the initiation of therapy and also from 63 healthy control participants. Approximately 50% of the patients in the study were also taking medications that affected the serotonin system. The resting connectivity analyses focused on cortical-striatal-thalamic circuits and their connectivity with other regulatory cortical regions such as the ventromedial prefrontal cortex (vmPFC). Cortical-striatal-thalamic circuits are of interest since they are thought to be involved in mediating habitual behavioral responses, and increased activity in these circuits has been associated with OCD. Other evidence links regions such as vmPFC to fear extinction, a process thought to be involved in the therapeutic efficacy of exposure response prevention therapy. The results of the clinical trial demonstrated that both interventions reduced symptoms and, as expected, the exposure response prevention group outperformed the stress management group. In relation to functional connectivity, relatively less vmPFC to left caudate and less vmPFC to thalamus resting functional connectivity were associated with better treatment responses in the exposure response prevention group, whereas the opposite relations were observed in the stress management group. When characterizing patterns of resting-state connectivity that predicted treatment response for participants in both groups, the researchers found that individual differences in connectivity between cingulo-opercular and frontoparietal regions with other subcortical regions predicted greater reductions in symptoms over the 12-week intervention period. Age-related effects were also found for the relation between resting functional connectivity and treatment response. For example, regardless of the type of treatment a patient received, in adolescents with OCD, increased connectivity between regions of the frontoparietal network with the nucleus accumbens predicted treatment response, whereas in OCD adults, increased connectivity between the frontoparietal network and dorsal putamen predicted treatment efficacy. It is interesting that while associations between resting functional connectivity and treatment response were found within the OCD group, when comparing the OCD group to controls no significant differences in resting functional connectivity were detected. In their editorial, Drs. Chadi Abadallah, Sameer Sheth, Eric Storch, and Wayne Goodman from Baylor University (14) argue that it is time to move beyond neuroimaging analytic strategies that are based on individual regions of interest like those used in this study to methods that are brain wide and multivariate in nature.
The articles in this issue of the Journal provide new insights into neurodevelopment-related psychiatric disorders such as ASD, ADHD, and OCD. Major findings reported in these papers include: 1) a lack of support for the reliable use of current biomarkers in ASD research; 2) recommendations, criteria, and empirical data aimed at assuring reliability in future ASD EEG biomarker development; 3) neurotypical male-like brain structure is associated with ASD in males and increased social-related cognitive deficits in ASD females; 4) rare pathological neurodevelopment-related genetic variants detected in a large health care population are associated with ASD, ADHD, schizophrenia, and bipolar disorder diagnoses; 5) the numerous genes associated with the risk to develop ADHD may also confer risk for the exposure to other ADHD risk factors; and 6) early evidence suggesting an association between resting-state functional connectivity measures and treatment outcomes in OCD patients receiving psychotherapy.
While ASD, ADHD, and OCD are the focus of this issue, it is important to keep in mind that neurodevelopment plays a role in most psychiatric illnesses, as genetics and early life experiences are broad determinants of risk. For this reason, and the fact that many of the illnesses that we treat in psychiatry emerge in youth or during early adulthood, it is reasonable to conceptualize these psychiatric disorders as having neurodevelopmental etiologies. Pursuing the neurodevelopmental mechanisms that underlie psychiatric disorders is important as a more complete understanding will inform the development of early intervention strategies that have the potential to prevent the life-long suffering associated with many psychiatric illnesses.
Department of Psychiatry, University of Wisconsin School of Medicine and Public Health, Madison.
Send correspondence to Dr. Kalin ([email protected]).
Disclosures of Editors’ financial relationships appear in the April 2022 issue of the Journal.
1.Parellada M, Andreu-Bernabeu A, Burdeus M, et al.: In search of biomarkers to guide interventions in autism spectrum disorder: a systematic review. Am J Psychiatry 2023; 180:23–40Link, Google Scholar
2.Webb SJ, Naples AJ, Levin AR, et al.: The Autism Biomarkers Consortium for Clinical Trials: initial evaluation of a battery of candidate EEG biomarkers. Am J Psychiatry. 2023; 180:41–49Link, Google Scholar
4.Autism Speaks: Autism Statistics and Facts, 2022. https://www.autismspeaks.org/autism-statistics-asd. Accessed Nov 15, 2022 Google Scholar
7.Shimelis H, Oetjens MT, Walsh LK, et al.: Prevalence and penetrance of rare pathogenic variants in neurodevelopmental psychiatric genes in a health care system population. Am J Psychiatry 2023; 180:65–72 Link, Google Scholar
8.Martin CL, Wain KE, Oetjens MT, et al.: Identification of neuropsychiatric copy number variants in a health care system population. JAMA Psychiatry 2020; 77:1276–1285Crossref, Medline, Google Scholar
9.Dewey FE, Murray MF, Overton JD, et al.: Distribution and clinical impact of functional variants in 50,726 whole-exome sequences from the DiscovEHR study. Science 2016; 354:aaf6814Crossref, Medline, Google Scholar
11.Brikell I, Wimberley T, Albiñana C, et al.: Interplay of ADHD polygenic liability with birth-related, somatic, and psychosocial factors in ADHD: a nationwide study. Am J Psychiatry 2023; 180:73–88Link, Google Scholar
13.Russman Block S, Norman LJ, Zhang X, et al.: Resting-state connectivity and response to psychotherapy treatment in adolescents and adults with OCD: a randomized clinical trial. Am J Psychiatry 2023; 180:89–99 Link, Google Scholar
14.Abdallah CG, Sheth SA, Storch EA, et al.: Brain imaging in psychiatry: time to move from regions of interest and interpretive analyses to connectomes and predictive modeling? Am J Psychiatry 2023; 180:17–19 Link, Google Scholar