Autism

Learning Objectives

  • Autism Spectrum Disorders (ASD) are significantly influenced by polygenic variation.
  • Smaller contributions come from heritable/de novo Copy Number Variations (CNV) and Single Nucleotide Variations (SNV).
  • The explicit role of specific environmental factors remains undefined.
  • The increasing global prevalence of ASD is largely attributed to improved identification, especially in individuals with a normal IQ range and in females.
  • Significant overlap exists in genetic risk across various neurodevelopmental disorders, including schizophrenia and ADHD.
  • The specific biological and environmental factors that lead to the dominance of one disorder over others are still unknown.

Early Diagnostic Criteria and the Evolution of Autism Definition

  • 1979: Mike Rutter documented the diagnostic "chaos" for autism in London.
  • 1982: This influenced the first formal definition of autism.
  • Rutter questioned the existence of autism subtypes and how to classify children showing some but not all features.
  • His diagnostic criteria were included in DSM-III (1980).
  • He also conducted the first genetic study of twin concordance in autism.
  • 1977: Susan Folstein & Michael Rutter indicated the importance of a genetic factor in the aetiology of autism.

Early Diagnostic Criteria (DSM III - 1980)

  • Michael Rutter, Eric Schopler, and Linda Lockyer outlined early diagnostic criteria:
    • Impaired social development, including a lack of social responsiveness.
    • Delayed and deviant language development if speech is present.
    • Bizarre responses to environmental stimuli and resistance to change.
    • Onset before 30 months of age.

Triad of Impairments (ICD-10, DSM-IV - 1992)

  • Reciprocal social interaction skills.
  • Communication, including non-verbal skills and imagination.
  • Repetitive, stereotyped behaviors, and inflexibility.

Dimensionality of Autism Symptoms

  • 2005: The autism triad was dimensionalized, revealing a dyadic pattern of symptom severity.

Current Conceptualization (DSM-5 - 2013, ICD-11 - 2022)

  • DSM-5: Social Communication Disorder
    • Social communication deficits.
    • Repetitive, stereotyped interests and behaviors, and sensory sensitivities.
  • Language delay is no longer a criterion.
  • Onset before 30 months is not specified.

Prevalence

  • 1980: 4 per 10,000
  • 1992: 5 per 10,000
  • 2013: 100 per 10,000

Measuring Autistic Traits in the General Population (20 years ago)

  • Researchers tried to measure autistic traits in the general population in the USA and the UK.
  • BJPsych, 2005, JAACAP, 2011, Elise Robinson.

Autism in England: Underdiagnosis Assessment

  • The Lancet Regional Health – Europe 2023;29.
  • Study assessing underdiagnosis in a population-based cohort using prospectively collected primary care data.

Prevalence of ASD in Schools - England

  • In 2021, the schoolchild population in England was 8.9 million.
  • 163,401 children attending mainstream schools had an ASD diagnosis and special educational needs.
  • 45,752 attended Special Schools (28% of total).
  • The prevalence of ASD is at least 1.8% in school-age children with autism as a primary need.
  • Prevalence in the USA: 1 in 54.
  • Prevalence in the UK: 1 in 56.
  • Of children with a primary diagnosis of ASD:
    • 4.9% have specific learning difficulties.
    • 8.8% have moderate learning difficulties.
    • 8.1% have severe or profound learning difficulties.
    • 78.2% do not have learning difficulties.
    • 1 in 5 (19.6%) receiving support were female.

Early Genetic Research Challenges

  • Early linkage studies using family members proved over-optimistic.
  • Expectation to narrow the search for autism susceptibility genes to a small number of genomic regions.
  • Positional cloning of susceptibility loci by linkage analysis may be a formidable task.

De Novo Copy Number Variations (CNVs) and Autism (2005)

  • 2007: Discovery that genome-wide copy number variations could predispose to autism.
  • Strong Association of De Novo Copy Number Mutations with Autism.
  • de novo: χ2\chi^2, P value
  • Simplex autism: 118, 12, 0.102, 0.0005
  • Multiplex autism: 77, 2, 0.026, 0.59, Multiplex/simplex 0.043
  • Simplex + multiplex: 195, 14, 0.072, 0.0035
  • Controls: 196, 2, 0.010

Common Inherited Genetic Variants (2006)

  • Increasing recognition that many relatively common inherited genetic variants (alleles) contribute additively to the overall risk of severe mental disorders.

Polygenic Risk and Specificity for Major Disorders

  • Polygenic Risk implies specificity for major disorders, including autism.
  • Shared molecular neuropathology across major psychiatric disorders parallels polygenic overlap.
  • Rare de novo variation, Common Genetic Variation, Intermediate Phenotypes, Altered Brain Gene Expression, Symptom Domains, Brain Circuit Dysfunction, Cognitive Impairment, Negative Symptoms, Positive Symptoms, Mood Disturbance, Reward Dysregulation, Clinical Syndrome.

Exome Sequencing and Neurodevelopmental Anomalies (2013)

  • In a study, 80% were children with ‘neurologic phenotypes’ including DD, ASD, and ID.
  • A 25% molecular diagnostic rate was achieved through EXOME SEQUENCING of children with neurodevelopmental anomalies.

Contribution of De Novo Coding Mutations to ASD (2014)

  • Whole exome sequencing has proven to be a powerful tool for understanding the genetic architecture of human disease.
  • 13% of de novo missense mutations and 43% of de novo likely gene-disrupting (LGD) mutations contribute to 12% and 9% of diagnoses, respectively.
  • Including copy number variants, coding de novo mutations contribute to about 30% of all simplex and 45% of female diagnoses.

Inherited and De Novo Genetic Risk for Autism

  • Autism Specturm Disorder (ASD).
  • multiplex families (2+ affected children).

Deciphering Developmental Disorders (DDD) Study - UK (2011)

  • Study of children with suspected Mendelian disorders.
  • Identified 94 genes enriched in damaging DNMs, including 14 that previously lacked compelling evidence of involvement in developmental disorders.
  • Estimate that 42% of our cohort carry pathogenic DNMs in coding sequences; approximately half of these DNMs disrupt gene function and the remainder result in altered protein function.
  • We estimate that developmental disorders caused by DNMs have an average prevalence of 1 in 213 to 1 in 448 births, depending on parental age.

Genomic Diagnosis of Rare Pediatric Disease (2023)

  • Just 5% were recorded as ASD, 87% with ID or DD

Summarizing Genetic and Environmental Risks for ASD (2019)

  • 2019 study finds heritability >80%
  • Heritability 52% (inherited variants). Environment 48%.
    • Genetics 62%, Environment 38.1%.
    • Inherited common variants 49.8%.
    • Inherited rare variants 2.6%.
    • de novo variants 9.5%.

Classes of Genetic Risk for Autism Diagnosis (2019)

  • Gene mutation.
  • CNV.
  • Polygenic risk.

Polygenic Risk for Social Communication Impairment (2016)

  • SNP heritability measured for clinically diagnosed ASD in PGC (USA) and iPSYCH (Denmark).
  • Genetic correlations between SCDC (population trait) and clinical samples ~25% of genetic influences on clinical ASD also influence SCDC scores.

IQ and Autistic Traits

  • Further studies confirm common risk variants for ASD are more prominent in those without ID/DD. Severe ID (encompassing many syndromes that confer high risk of ASD) show far less heritability than that observed for mild ID intelligence in general, and ASDs.

Rare Coding Variants (2020)

  • Rare coding variants could be parsed into two broad groups
    • ASD predominant.
    • ASD + neurodevelopmental delay.
  • Prenatal gene expression regulatory gene (GER) anomalies lead to ASD + neurodevelopmental delay.
  • Postnatal neuronal communication (NC) genes lead to ASD predominant phenotypes.

IMAGINE Families

  • Cohort was similar in socioeconomic distribution and ethnic composition to the UK population.

IMAGINE Study (2015-2025)

  • Collaboration between UCL, Cardiff University & Cambridge University.

Neuropsychiatric Risk in Children with Intellectual Disability (2022)

  • Children with intellectual disability frequently have multiple co-morbid neuropsychiatric conditions and poor physical health.

Genotype-phenotype Associations in Children with CNVs (2019)

  • Copy number variations – 13 deletions/duplications.
  • Children aged 4 years and older with pathogenic CNV or single nucleotide variants were recruited via the UK National Health Service (NHS) medical genetic clinic network and via patient support groups to complete broad online phenotyping
  • We systematically assessed the children for psychiatric disorders and broader traits of neurodevelopmental, cognitive, and psychopathological origin and compared results of ND-CNV carriers with control siblings to test the hypothesis that phenotypes would differ by genotype, both quantitatively in terms of severity and qualitatively in the pattern of associated impairments.

Chawner et al Findings

  • The specific genotype accounted for a low proportion of variance in cognitive and behavioural outcomes (approximately 5–20% depending on the trait).
  • The 13 ND-CNVs studied have a similar range of adverse effects on childhood neurodevelopment, maximal in the domains of Autistic and ADHD traits.
  • Genomic risk for neuropsychiatric disorder has pleiotropic effects on multiple processes and neural circuits
  • Future research should avoid being narrowly focused on single phenotypes.

SNV focus in IMAGINE-ID cohort

  • All SNV evidence supported by Level A evidence within the IMAGINE ID cohort were identified.
  • Includes 204 unique genes, 119 genes on X-chromosome.

Autism Specific Genes in IMAGINE-ID cohort

  • Of these 36 genes, 7 are present within the IMAGINE cohort – highlighted in pink here. Dark pink =ASD + ID Light pink = ID only.

SNV - gene function

  • Translation regulation, Cilliopathy, Cell division, Cytoskeleton, RNA processing, Neurogenesis, Post-translational modification, Cellular process, Synapse, Intracellular signalling, Transcription regulation, Chromatin.

Association between gene function disrupted and prevalence of autism

  • Cellular process, Cell division, Transcription regulation, Post-translational modification, Intracellular signalling, Neurogenesis, Synapse, Cilliopathy, Cytoskeleton, Chromatin, Translation regulation, RNA processing

Impact of loss of specific gene function on social communication skills

  • Synaptic dysfunction, Impaired intracellular signaling. Higher scores imply more impairment
  • LOW risk, HIGH risk

Impact of loss of gene function on repetitive and stereotyped behaviours

  • Synaptic dysfunction, Disrupted RNA processing. Higher scores imply more impairment. Higher scores imply more impairment
  • LOW risk, HIGH risk

Take-Home Messages

  • Since 1979, diagnostic criteria for autism have shifted substantially, and apparent prevalence has substantially increased.
  • Diagnostic ‘chaos’ has been reduced by standardizing criteria, but have we therefore lost sight of diversity, the many possible autisms?
  • Why does a ‘genetics-first’ approach to genetic risk fail to replicate the findings of ‘phenotype-first’ studies for so many autism genes?
  • Could we gain greater insight into the biological mechanisms underlying ASDs by paying more attention to phenotypic variants?