genes & environment in autism

Question 1

evidence for genetic involvement in autism

Answer 1

• higher co-occurrence of autism in MZ twins than DZ twins
• 1st degree relatives of autistic people have an increase in behavioural/cognitive features associated with autism
  –> the broader autism phenotype
• autism often co-occurs genetic conditions
  –> e.g. 16p11.2
  –. e.g. Sotos syndrome

Question 2

twin studies in autism

Answer 2

• first twin study of autism was in 1997
• Hallmayer et al., (2011)
  –> concordance rate of up to 77% in Mz twins
  –> concordance rate of 31% in Dz twins
  –> suggests there is a genetic link for autism but still don’t know what that is

Question 3

autism as a spectrum

Answer 3

• not categorically yes or no
• there is a spectrum
• we are all on it to some extent, some people are towards the more extreme end of the spectrum

Question 4

the broad autism phenotype

Answer 4

• behavioural features associated with autism investigated in parents of
  –> multi-incidence of autism families (MIAF)
  –> single incidence of autism families (SIAF)
  –> families with children with Down’s syndrome
• MIAF and SIAF showed an increase in autism-related features
• considered to provide evidence for ‘Broad Autism Phenotype’
  –> and also evidence for genetic origin of autism

Question 5

Sotos syndrome

Answer 5

• “overgrowth” syndrome
• prevalence = 1 in 14,000
• deletions on chromosome 5 involving the NSD1 gene
• diagnostic criteria:
  –> overgrowth with advanced bone age
  –> macrocephaly
  –> characteristic facial appearance
  –> intellectual disability

Question 6

Lane et al (2017) - Sotos syndrome

Answer 6

• recruited 78 individuals with Sotos syndrome
• measured traits of autism using the ‘Social Responsiveness Scale’
• found that 83% of participants scored above clinical cut-off for autism

Question 7

co-occurrence of autism and other genetic conditions

Answer 7

• autism occurs in neurodevelopmental conditions with known genetic origin (e.g. Sotos syndrome, Fragile X syndrome, 16p11.2)
• highlights that genetic alterations are associated with autism symptoms, providing support for the argument that there is genetic association to autism
• some genes are more linked with autism symptoms than others

Question 8

complex genetics

Answer 8

• involves a large number of different genes, and most likely interactions between different genes
• approx 65 different genes have been strongly linked to autism
  –> genes are most involved in brain development
• single genetic cause for autism is not known and very unlikely to exist
  –> appears there are many genetic routes to autism
• only 10%-20% of autism cases are accounted for by known genetic abnormality

Question 9

environmental epidemiology

Answer 9

observation of human populations to infer the environmental causes of a condition or disease

Question 10

environmental links to autism

Answer 10

• important to avoid unscientific reports in the media
• important to carry out good, rigorous science in an attempt to understand potential environmental causes of autism
• researchers recognise that environmental factors may contribute, in part, to the growing incidence of autism

Question 11

example of bad science

Answer 11

• 1998, Wakefield et al. claimed that there was a link between the MMR vaccine
• paper was based on a very small sample of 12 children
• led to a media frenzy and to 1000s of people across the glob choosing not to vaccinate their children
• consequently more people are now contracting measles across the world

Question 12

impact of Wakefield et al.’s work

Answer 12

• global increase in measles.
• lack of trust around research on environmental causes of autism

Question 13

does the MMR vaccine cause autism?

Answer 13

• paper was later retracted as there was evidence that parts of the work were fraudulent
• many follow up studies have since found no evidence for a link between the MMR vaccine and autism
• in 2019, a paper found no evidence for a link between MMR and autism, in a population sample of over 650,000 children

Question 14

environmental factors linked to autism

Answer 14

• prenatal viral infection (e.g. influenza or rubella)
  –> effects are very dependent however on the immune status of the mother
• maternal diabetes
• maternal obesity
• prenatal stress
• teratogens (e.g. valproic acid for epilepsy or psychoactive drugs)
• pesticide exposure during first 8 weeks of pregnancy
• parental age (advanced maternal age)
• air pollution

Question 15

causation vs correlation

Answer 15

• incredibly tricky to identify clear environmental causes
• analysing environmental influences = correlations
• correlation does not equal causation
• drawing a direct line between cause and effect in a messy, complex world is inherently challenging

Question 16

how can we try to establish causality? - Hill’s criteria

Answer 16

• strength
• temporality
• plausibility
• consistency
• biological gradient
• coherence
• experiment
• specificity
• analogy

Question 17

environmental epidemiology

Answer 17

• environmental research is observational (not fully controlled)
• extraneous variables are also recorded
• results of one study are easily challenged
• how can we be confident in the results of our study?
  –> meta-analyses
  –> take numerous independent studies and combine their results to see if the relationship remains

Question 18

evidence for folic acid in pregnancy and risk of autism

Answer 18

• Suren et al. (2013)
  –> sample of 85,176 children
  –> folic acid during pregnancy reduced risk of autism
• Virk et al (2016)
  –> sample of 38,035 mothers
  –> no evidence that folic acid reduces the risk of autism

Question 19

air pollution: meta analysis (Larn et al., 2016)

Answer 19

• initial set of papers are screened to identify relevant work that fits the inclusion criteria
• papers are whittled down until only the ones used are highly specific, trusted and original
• combined data shows a small but significant association between particulate matter (part of air pollution) and autism

Question 20

Odds Ratios (OR)

Answer 20

• Odds Ratio provides a statistical method for understanding the association between exposure and an outcome
• ‘OR = 1’ means exposure to the cause does not lead to an effect
• ‘OR < 1’ means greater exposure to the cause is associated with LESS of an effect (good thing)
• ‘OR > 1’ means that greater exposure to the cause is associated with a LARGER effect (bad thing)

Question 21

Odds Ratios and risk of autism from MMR

Answer 21

• ALL datasets showed odds ratio < 1
  –> i.e. no association between MMR and autism

Question 22

do environmental factors autism?

Answer 22

• measuring environmental exposure accurately is very difficult
• very little clear evidence for environmental causes of autism
• possible suggestions focus on particulate matter (air pollution) as a risk factor
• however evidence for a direct causal link between environmental factors and autism is generally weak

Question 23

gene x environment interaction

Answer 23

genetic risk + environmental risk = increased likelihood of developmental condition (e.g. autism)

Question 24

gene x environment interaction and autism

Answer 24

• some evidence for involvement of air pollution in autism risk
• one (of the many) genes that has been associated with increased risk of developing autism is the MET gene
  –> the MET gene codes for proteins that trigger mitogenesis (cell division) and morphogenesis (structural development)
• in animal models, air pollution has a direct effect on MET expression

Question 25

gene x environment interaction and autism summary

Answer 25

• link between air pollution and autism risk
• link between air pollution and MET gene
• link between autism risk and MET gene

Question 26

volk et al. (2014) - method

Answer 26

• 252 autistic children, 156 neurotypical controls recruited from Californian preschools
• MET genotype was assessed via blood test
• exposure to air pollution was measured / estimated for each participant

Question 27

volk et al. (2014) - results

Answer 27

• interaction effect:
  –> children with both an alteration to the MET gene AND increased exposure to air pollution were at increased risk of autism than those with just the MET alteration, or just those with increased exposure to air pollution
  –> presence of both has greater impact than either factor alone

Question 28

summary

Answer 28

• genetic Basis to autism
  –> twin studies
  –> Broader Autism Phenotype
  –> occurrence of autism in developmental disorders with known genetic cause
• environmental Basis to autism
  –> difficult to study
  –> some evidence for links
  –> potentially more fruitful to look for gene x environment interactions
• value of meta-analyses
• understanding of how to interpret Odds Ratios in research