2


Babies and infants

Age of identification
‘The earlier intervention can begin, the better the outcome.’1

Data from both the USA (e.g. California2) and Europe (e.g. Denmark3), show
that the average and median ages at which childhood autism diagnoses are made
are steadily dropping. However, there are some subtleties in the pattern; for
example, our analysis4 suggested that in England the average age of diagnosis for
the youngest children (0–​2 years) went up marginally between 1998 and 2018,
perhaps because of increased demand for diagnoses and long waiting times.
Claims for identification of pre-​symptomatic predictors of autism are now
being made for very young babies.5, 6 Using brain imaging, one group has noted
autism-​specific ‘features’ in six-​month-​old babies,7 while another, which received
world-​ wide media attention, used eye-​ tracking technology to identify subtle
differences in the way affected babies responded to visual prompts:8 ‘Autism can
be identified in babies as young as two months, early research suggests’.9 Studies
such as these, and others, are used to define infants ‘at risk’ of autism. To be ‘at
risk’ is to be in danger of falling outside the statistical norm –​a state requiring
expert advice, intervention, parental regulation and surveillance.
The narrative of earlier-​is-​better (EIB) transcends autism to pervade child
psychiatry, education and infant development and beyond: dementia, dia-
betes and hypertension spring to mind as examples of ways in which medicine
has extended its jurisdiction.10 Identifying potential early signs and signals of
autism makes earlier diagnosis, detection and intervention possible. However,
although the evidence base is regularly reviewed, the evidence that earlier inter-
vention results in more successful outcomes for the child is poor.11, 12 A recent
UK review of evidence on screening infants for autism, conducted in 2011,
concluded:

• Diagnoses of very young children may not be stable.

• Current screening tools are insufficiently sensitive and may not be accepted
by a significant proportion of parents.
• The outcomes of interventions are variable.
• It is not known if short-​term improvements continue in the long term.13
32 ‘Artefactual’

Baby-​sibs
The collection of studies known as ‘baby-​sibs’ research gives ‘at-​risk’ status to
new-​born and unborn children who have siblings with autism.14 At-​risk status
is given because autism is heritable and geneticised.15–​18 Sharing a genomic pro-
file with autistic siblings, that is, being a sibling of someone with autism, there-
fore puts you at risk of autism. Estimates of the extent of familial heritability
over 40 years ago were that around 90% of variance in autistic traits is attribut-
able to inherited factors,19 whereas today around 50% of variance is attributed to
inherited factors.20
These two types of identification of the youngest children (an autism diag-
nosis in babyhood and the ‘at-​ risk’ status given to babies and unborn chil-
dren) are related but distinct processes. Earlier autism diagnosis has consistently
been associated with more severe autism and more severe impairment.21–​26 In
childhood studies, the factors associated with an earlier diagnosis include greater
language delay, need for a greater degree of support, more cognitive and intel-
lectual disability, greater parental concern, an autism (as opposed to Asperger’s)
diagnosis and the severity of autistic beahviours.21–​26 The picture is one in which
more severe autism is more obvious, therefore is picked up earlier in a child’s life.
Put simply, babies with more extreme neurodevelopmental difference are, and
were before 1990, easier to spot.
A raft of EIB studies tells the story of how the earlier a child can be recognised,
the more effective early intervention is, and so it must be brought into place.27–​30
The longer diagnosis is delayed, the greater the chances of missing a critical
developmental period.22 Once this window is missed, brain plasticity is lost and
interventions may be ineffective.
At-​risk babies (such as baby siblings, through their shared inheritance of a
genetic predisposition) may be anywhere in the broad autism phenotype, which
includes sub-​clinical (milder) levels of autistic traits.31 Baby-​sibs studies look for
early indicators of autism but necessarily include children who go on to develop
milder, and in some cases, few or zero, autistic traits. Many but not all, baby-​sibs
studies follow up on later autism diagnosis.
Precursor signs of autism in infants, which have been deduced from baby-​
sibs and retrospective studies, can be loosely divided into behavioural signs,
genetic predisposition and neurological differences. Behaviours include types
of movements or lack of motor skills, imitation impairments, lack of physical
exploration of objects in the environment with less object manipulation32 and
lack of joint attention. Many studies identify abnormal movement as a pre-
cursor, including gross motor, fine motor and postural control28, 33–​36 and babies’
head lag.37
The larger catchment of ‘at risk’ of, as opposed to diagnosed with, autism
presumably results in some studies widening the net of potential early signs of
autism gleaned from siblings’ behaviours and abilities. An aspect that is not often
dwelled on is that a researcher denoting a baby sibling as ‘at risk’ surely makes
an autism diagnosis more likely, not only through relatedness but also if parents
 Babies and infants 33
see their baby as being in a proto-​autism group. This, one would assume, will
increase the likelihood of referral to a clinic and, once in the clinic, the interpret-
ation of behaviour as autism. At the same time, by defining new signs of precursor
autism using behaviours in the ‘at-​risk’ group, all ‘at-​risk’ babies’ behaviours start
to be understood as signalling autism. It seems circular: what counts as a specific
‘signifier’ of autism, most often motor difficulties, becomes connected to identi-
fication of autism in the group from whom the ‘signifier’ was determined.

Earlier diagnosis and risk

As well as identifying early indicators, autism studies have early diagnosis of
autism as a core objective.38 EIB is most commonly operationalised in interven-
tion research. The assumption is that there is a fixed disorder that is present from
birth, can be correctly identified soon after birth and which intervention will
ameliorate. Advocacy, funding and charity organisations also strongly promote
earlier diagnosis; for example, Autistica’s report, One in a Hundred, emphasises
the importance of diagnosis at the youngest possible age.39 This report is typ-
ical of policy guidelines in higher-​income countries but the rhetoric of early
diagnosis is also visible in narratives aimed at broader publics. In the USA, five
million coffee cups were released by Starbucks in a campaign aimed at raising
the profile of autism, put together by the founder of the charity Autism Speaks
(Figure 2.1).
In an inspired analysis, Anne McGuire argued the Coffee Cup casts the non-​
normatively developing child as non-​valuable and perhaps even non-​viable in a
market-​driven economy (of Starbucks).40 Certainly, this widely distributed dec-
laration contributed to the cultural recognition of autism as a threat, something
to be dreaded and something to be identified (by parents’ surveillance) as early
as possible so that it can be fixed. And the younger the better. It also invokes a
moral obligation for parents to monitor their children, if they wish to qualify as
good parents.
The Coffee Cup uses non-​gender-​specific language. Despite this, it is inter-
pretable as an exhortation to good mothering. The word ‘parent’ is gender-​blind
and obliterates oppressive imbalances in the roles and experience of mothers by

Figure 2.1 The Coffee Cup example: Starbucks Autism Awareness campaign.

34 ‘Artefactual’
using the gender-​neutral language of ‘parenting’.41 In autism discourse, ‘parent’
is often a synonym for ‘mother’, because the vast majority of primary carers of
autistic children are mothers. Studies of parental attitudes, parent-​rated behaviour
scales and parent-​mediated interventions often overwhelmingly rely on mothers
to participate. Good mothering tacitly means offering as much therapy as possible
to the child, at the expense of any other career; Gil Eyal and colleagues42 refers to
this as the ‘vocation’ of autism parenting.
The threat of autism, this framing of risk, prompts anxiety which demands
action. EIB targets the family, in partnership with medical institutions, as the
site of early detection and intervention. Intervention may involve one-​to-​one
teaching or up to 40 hours of speech, occupational and Applied Behavioural
Analysis (ABA) therapies a week.42 One US survey suggested parents use as many
as 111 different therapies;43 the mean number used at any one time was seven. The
more severe the autism, the more types of treatments parents experimented with.
In her work on attention deficit hyperactivity disorder (ADHD), Singh situates
mothers’ actions in the context of the multiple pressures they feel from so many
sources, such as the Coffee Cup campaign.44 A patriarchal culture that allows
mothers to be culpable of their children’s behaviour, responsible for monitoring
the progress of the child and for ‘doing something’ if autism is detected, is a
driver for the adoption of highly suspect therapies. The neurodiversity movement
(Chapter 4) encourages better choices by situating autism as non-​problematic –​a
condition that cannot, and perhaps should not, be ‘fixed’.
The moral obligation for mothers to treat, monitor and report to clinicians
is not new;45 through a sociological lens, it is a form of surveillance medicine.46,
47
Surveillance medicine, the screening, monitoring and establishment of early
risk factors, involves monitoring across a whole population, including healthy
people.46 Sociologists such as Ulrich Beck have pointed to a ‘politics of anxiety’ in
the risk society.48 David Armstrong writes about how infants were the first popu-
lation to be scrutinised and surveyed for potential risks to normal childhood, such
as being of a height and weight that fall outside statistical norms.46
Concepts of surveillance draw on Michel Foucault’s work, particularly his
book Discipline and Punish,49 in which he describes how people are monitored,
understood and regulated via institutions, which for babies include nurseries,
research institutes, health visits and baby clinics. Foucault describes how people
are first trained and observed in institutional settings to produce knowledge
about disciplinary norms (for example, the observation of babies in maternity
hospitals that produces knowledge about paediatrics, or the knowledge produc-
tion of baby-​sibs studies) and subsequently populations become monitored and
subject to regulatory controls. Screening and surveillance therefore promote
framing and recognition of differences as problems that were formerly not part of
a medical remit. Hence, for good or ill, surveillance fosters medicalisation. The
community is encouraged to monitor others in the community, providing nor-
mative standards of behaviour. This community policing and neighbourly surveil-
lance were heightened during the 2020 Covid-​19 lockdown to maintain social
and behavioural norms.
 Babies and infants 35
Foucault wrote about the historical steps from a past model of external
monitoring and top-​down surveillance by powerful actors in the establishment,
such as monarchs and lawyers, to community surveillance that provides a net-​like
power structure in which everyone is responsible for upholding normal behav-
iour, to the inculcation of internalised self-​surveillance, the internalisation of
bio-​power, so that one comes to ‘subjectivise’ oneself and discipline one’s own
body.49 Mothers’ internalisation of vigilance and responsibility for the monitoring
of their child seem to be an example of a relational form of bio-​power.
In the case of the Coffee Cup, an exhortation for parents to perform sur-
veillance of their child invokes anxiety, with autism described like a threatening
disease. The Coffee Cup therefore promotes both pathologisation and vigilance
and invokes autism as an object in itself, distant and removed from the individual
person, meaning a person may become alienated from it.50 For Foucault, a condi-
tion such as autism is objectified or ‘spatialised’ by its description as an entity that
exists independently of the person (in texts and on coffee cups, etc.). Diagnosis
locates autism in a second space, the brain, but autism also requires a third space,
the social realm, because it is rendered in interaction. According to Foucault,
‘truth’ is produced through these levels of spatialisation, exercised by the profes-
sional gaze.51 Once objectified, autism (or any condition) is subject to discipline,
and through its control, subjection leads to the subjectification of people who are
diagnosed. Although young children may not be able to resist this, adults can –​a
topic I will return to in Chapter 4. But Foucault was a master of the rhetorical
device; others see power dynamics very differently, with less sinister overtones.
A similar rhetorical device to that of the Coffee Cup (risk, threat, requiring
action) appears in most medical funding applications that try to identify early signs
of autism. Research into either biomarkers or behavioural markers in infancy usu-
ally starts with a statement about autism’s terrible impact on personal outcomes,
families and the economy. Autism is often positioned as an object that is thor-
oughly bad news, the threat of which provokes anxiety and should be eliminated
as early as possible.
Selective interpretation of data justifies the use of language to back up the
EIB story, such as Green and colleagues’ study of intervention for at-​risk babies
in which parents delivered the intervention.52 Results were described as ‘encour-
aging’ despite there being no significant improvement in the primary outcome
(attentiveness to parent); indeed, a few babies had a worse outcome. The abstract
describes first how ‘point estimates suggest the intervention increased the primary
outcome of infant attentiveness’, although qualifies this as ‘including possibilities
ranging from a small negative treatment effect to a strongly positive treatment
effect’ (actually it had a non-​significant effect). The positioning and wording of
reporting, in this and other literature, bolster the EIB narrative by accentuating
the positive and diminishing the negative of EIB. Green and colleagues correctly
reported the possibility of a negative outcome but the results were nevertheless
framed as ‘exciting’ in the promise of intervention research.
Another example is a research paper, published in the journal Autism,
that analysed the socio-​demographic and child-​based factors that predict late
36 ‘Artefactual’
diagnosis.26 The discussion describes how children are at risk of late diagnosis
(after five years old): ‘our understanding of “red flags” for missed diagnosis, that
is early characteristics for children at risk of receiving a late diagnosis’ (my italics).
The phrase ‘red flags’ indicates autism is something that should raise an alarm and
being ‘at risk’ of receiving a late diagnosis is troubling.
The Coffee Cup, and other forms of the EIB narrative, exhorts parents (specif-
ically mothers) to perform surveillance and early childhood monitoring, to report
proto-​autism behaviours and, if possible, to intervene early. This surely leads to
more early referrals and ultimately more diagnoses, contributing, perhaps in a
small way, to autism’s rise.

Caveats to EIB
There is a lack of evidence that diagnosis is stable at younger ages.13 At very
young ages, it is difficult to distinguish an autistic from an allistic (non-​autistic)
child, to distinguish a toddler who is not speaking because they may continue to
display traits of autism later in life from a toddler who is a slow developer and will
catch up. Some children grow out of autistic traits: 30% of children who are given
a diagnosis at two years old no longer meet the criteria for an autism spectrum
disorder (ASD) diagnosis at four.53
There is more uncertainty about future trajectories when screening procedures
for autism begin before the child is two.54 Our work followed the trajectories of
two groups of children from two years old to 12; both groups were measured
with comparably severe autistic-​type traits at age two. The children in one group
received an autism diagnosis, while those in the other did not.55, 56 At adolescence,
the children without an autism diagnosis were better on a range of outcomes. In
other words, some pre-​school-​age children who have autistic traits can improve
to sub-​clinical levels without having ever been diagnosed or treated. In these
cases, ‘wait and see’ may indeed be the best strategy.
To me, our work underlined that the human child is born in an immature
state and learns adaptive behaviours as they grow. Many behaviours characteristic
of developmental disorders are noticeable in all younger children: hand flapping,
hyperactivity, inattention and motor difficulties are all common in toddlerhood.
Resolving, at a very early stage, who has a lifelong impairment (and what impair-
ment) and who will catch up is extremely difficult. In medical parlance, the spe-
cificity of these early signs in predicting autism may be very low, with many false
positives. In a prospective Danish cohort of more than 75,000 children, in infancy
the signs that distinguished autism from intellectual disability were unclear and at
18 months old, the positive predictive values (the probability that subjects with
a positive test truly have autism) were below 10% for both individual predictors
and aggregated risk scores.57
In addition, as children grow up the extent of autistic behaviours tends to
diminish.58 The age effect is illustrated by the seasonal influence on ADHD diag-
nosis. Summer-​born children are more likely to be diagnosed with ADHD; a sys-
tematic review showed ADHD is consistently diagnosed more often in children
 Babies and infants 37
who are young for their school year (which starts in the autumn in the UK),59
not because they have more ADHD but simply because, relative to their peers,
younger children display more behavioural characteristics of ADHD. Taking a
developmental perspective therefore throws up challenges to the current recom-
mendation for the reduction of age of diagnosis of autism to very young children.
Another counter to EIB is that diagnosis is not a neutral process of identifica-
tion but shapes how others react to the baby. Given a specific childhood diagnosis,
the people around the child (parents, teachers and clinicians), tend to interpret the
child’s behaviour in the diagnostic frame.60 This may lead to an expectancy bias, in
the classroom for example, that negatively affects outcomes.61–​63 Thus, very early
labelling is problematic even if you consider a young baby either categorically has
autism or does not, which is debatable. If the diagnosis is a false positive, those
around the child might look at them through an autism lens; could this not nega-
tively affect their trajectory?
Advocates of early diagnosis, on the other hand, see early identification as
a crucial step to enable access to support and accommodations that benefit all
children; diagnosis opens the gateways to intervention.64 Autism can certainly
act as an explanatory frame for differences in a child’s biological and psycho-
logical make-​up, which can radically improve the functioning of the family. As we
have seen, autism researchers have emphasised the critical importance of inter-
vening early in autistic children’s lives to give them the best chance of meaningful
communication.
A final caveat is that, despite the overwhelming call for early intervention, sys-
tematic reviews suggest research into early interventions is of poor quality and the
effectiveness of early intervention is not proven for children with autism.12 The
rhetoric around early identification is widespread, and therefore should be under-
pinned by a rigorous evidence base. In fact, randomised controlled trials (RCTs)
on early interventions are rare. One systematic review uncovered a replicated
finding that many children who receive early intensive intervention, across meth-
odologies, do not demonstrate dramatic gains in social, cognitive, adaptive and
educational functioning or autism-​specific behaviours.12 A more recent review on
the effects of ABA concluded there is weak or very weak evidence that ABA is a
useful behavioural treatment for some children with autism and none that it alters
core autism symptoms.65
The best that can be concluded is that some interventions improve some areas
of functioning and sometimes improve cognition, in some young autistic chil-
dren, some of the time. What is not often acknowledged is that early interventions
for autism have high costs both for the children and in terms of parents’ financial
and time commitment. Programmes involving more than 40 hours of intensive
therapy a week may be exhausting for parents (disproportionately mothers) and
children alike.42 The extra parenting work (usually mothering work) is implicitly
expected to be done at home, even though a better outcome is not guaranteed.
Nor is it currently possible for a clinician to confidently recommend a particular
treatment for a particular child. There seems to be a disjunction between the
level of actual evidence for the efficacy of early interventions for autism and what
38 ‘Artefactual’
I would term the rhetoric of early intervention and surveillance that designates
good mothering.

Biomarkers
‘At-​risk’ status can also be assigned from the evidence of biomarkers: objective,
biological, measurable differences. For some conditions, biomarkers are physical
attributes such as weight or heart rate; for autism, the biomarkers are usually
neurological or genetic differences.
Some researchers use indices of risk or algorithms that calculate from a com-
bination of biomarkers. For example, for ADHD, a genetic risk profile combines
a number of genetic markers into an overall at-​risk-​of-​ADHD score, a poly-
genic risk score.66 In this way, researchers increase the predictive power of their
models and, based on a risk index, can calculate a person’s estimated probability
of developing a condition. Considering an at-​risk group in this way often gives
access to larger and younger populations than would be possible if only confirmed
cases were considered.
Perhaps the ultimate in baby surveillance is an electronic romper suit that
monitors all aspects of the wearer’s behaviour for ‘warning’ signs. In 2015,
I interviewed a technology expert with many years’ experience of designing com-
puter algorithms to detect mouse movements in the laboratory. He described
his company’s on-​going project to design romper suits to be used in the home
to recognise the autism behavioural phenotype and help detect autism.67 This
‘smart’ baby suit has sensors woven into the fabric that monitor the baby’s heart
rate, respiration, mobility and movement against normal parameters and auto-
matically and securely transmit the data to the researchers’ lab. The design was
commissioned by at-​risk-​of-​autism researchers but perhaps will be rolled out to
the general population. Late development and missed milestones will ring ‘alarm
bells’, raise ‘red flags’ and provide the required ‘early warnings’.
For many years, there has been a push to detect biomarkers of autism because,
as some argue, a biomarker is considered to be a more objective measure, and
potentially a better mechanism of identification, than behavioural clinical
assessments68 which are subjective and dependent on the settings in which they
are recorded. Plausible biomarkers for autism are measures such as brain cir-
cumference, genetic profile or a particular pattern of activity in the brain during
a certain task, normally revealed by magnetic resonance imaging (MRI), but
studies have identified many others.69 Some scientists have advocated the fusing
of behavioural definitions with biological, particularly neurological, indicators,
for all psychiatric classes.70 Perhaps, neither is ‘better’; just different. Publicising,
operationalising or adjusting either definitions or indicators will both influence
our understanding of the autism category and alter who is in it. Autism is partly
a product of how it is measured and identified.
In medical discourse, ethical arguments regarding ‘at-​risk’ status are often
founded on the notion of false positives. Statements about diagnosis and at-​risk
status use terms such as inaccuracy, misdiagnosis, false positive and validity. In this
 Babies and infants 39
language, being named at-​risk-​of-​autism may not accurately reflect your status
(you may be a false positive), leading to misdiagnosis and raising questions about
the validity of the at-​risk category. These terms confer notions of ‘truth’, ‘fact’
and an ‘objectivity’ to be striven for. Again, these words assume there is a true
fixed autism to be measured against and that conferring at-​risk status does not in
itself shape how we understand and define autism, how often we refer for autism
and how deeming babies at risk could alter their developmental trajectory.
To take an example, let’s say a neuro-​ marker is discovered, for example
differences in white-​matter tracts,69 that forms a biomarker to identify autism.
The at-​risk group of babies thus identified will be a slightly different bunch to the
babies identified as at risk by their behaviours, such as head lag. In this hypothet-
ical example, publicising the neuro-​work leads to understandings of autism as a
neural condition (the white-​matter tract difference). Atypical white-​matter tract
at-​risk babies are more likely to be referred and diagnosed. Thus, the net effect
of finding biomarkers contributes to what being ‘at risk’ of autism looks like, and
who qualifies as having autism may be very slightly reshaped.
Although biomarker results are frequently described as ‘promising’, they are
not often replicated or applicable to the whole spectrum. However, the search
for genetic markers for autism has revealed some very useful markers of rare
syndromes, for example Williams syndrome. The genetics of autism are complex,
with different genetic sub-​profiles that involve multi-​faceted interactions with the
environment.71–​73 Because what is diagnosable as autism is a slowly moving target,
the search for a fixed set of biomarkers against which to compare is like having
moving goalposts; it may be better to search for sub-​groups across the spectrum.
The latest iteration of the Diagnostic and Statistical Manual of Mental Disorders
(DSM-​5) has dropped the distinction between Asperger’s disorder and autistic
disorder but acknowledges differences within the autism spectrum, which is now
stratified by the severity both of social communication impairment and restricted
and repetitive patterns of behaviour, and with and without co-​occurring intellec-
tual disability. In DSM-​5 the autism spectrum is also codified by known genetic
conditions, biomarkers, although only a small percentage of cases have known
genetic markers.
Despite the move towards sub-​grouping, there is still investment in discovering
the genetics of autism across the whole spectrum. Some research groups aspire to
create a genetic test for autism that could be administered before birth, and some
commercial laboratories offer parents a non-​invasive pre-​natal test they claim can
screen for mutations in a range of genes, including some related to autism.74 This
claim has provoked an outraged reaction from the autistic community. In 2005,
the autistic activist Meg Evans created the Autistic Genocide Clock as part of her
Star Trek fanfiction website, Ventura33. Evans became mobilised after joining
the autistic forum, Aspergia, and later the chatroom Aspies for Freedom, founded
by Amy and Gareth Nelson, who also published a declaration that autistic people
should be recognised as a minority group.75 The Autism Genocide Clock was a
ten-​year countdown in the image of a clock; it responded to and resisted a pro-
nouncement in 2005 that genetic research on autism could lead to a genetic test
40 ‘Artefactual’
within ten years. Evans’s point was that a pre-​natal genetic test for autism could
lead to abortions of foetuses that test positive for autism –​in her view, a form of
genocide. Writing in a collection of stories about autistic activists released as part
of our Exploring Diagnosis project,76 she described her timer clock as a reaction
to autism discourse that, as she puts it, says ‘the world should not have autistic
people in it’. Evans took the clock down in 2011.
Certainly, the work towards pre-​natal testing positions autism as a suitable
rationale for abortion. Presumably such a test would be accompanied by genetic
counselling for parents who chose to take it, to support them to decide whether
to abort a baby with autism. Having been through such a scenario myself (when
I was pregnant, my daughter screened positive for being at risk of Edwards’s
syndrome; it turned out to be a false positive), I know both how stressful this
process can be for parents, and how powerful and potentially life changing the
medical concepts can be in practice.
Evans’s argument parallels those made by members of the disability rights
movement, that pre-​natal genetic tests are a form of eugenics, leading towards
the elimination of people like them, and that allowing abortion on the grounds of
disability is discriminatory.77 Others argue quality of life is important to consider.
Edward’s syndrome leaves babies with heart, respiratory, kidney and gastrointes-
tinal conditions, with 87% dying before one year old. The Autistic Genocide
Clock illustrated the tension between a newer progressive, affirmative model of
autism-as-identity and an older model of severe autism with co-morbidity and
complications in a medical frame. Evans’s strong language has parallels with his-
toric resistance to the elimination of other minority groups.78
The twin processes of pushing back age of diagnosis into infancy and defining
infants as ‘at-risk’ may have both contributed to the rise in autism observed in
Chapter 1, if in a minor way. Earlier diagnosis contributes directly as a younger
cohort is eligible for diagnosis. ‘At -risk’ status may contribute indirectly through
widening ‘what counts’ as autism. Yet a more seismic shift in diagnostic practice
occurred at the life stage covered in the next chapter: childhood.

References
1. Woods, J. J. & Wetherby, A. M. Early Identification of and Intervention for Infants
and Toddlers Who Are at Risk for Autism Spectrum Disorder. Lang. Speech Hear. Serv.
Sch. 34, 180–​193 (2003).
2. King, M. D., Fountain, C., Dakhlallah, D. & Bearman, P. S. Estimated Autism Risk
and Older Reproductive Age. Am. J. Public Health 99, 1673–​1679 (2009).
3. Parner, E. T., Schendel, D. E. & Thorsen, P. Autism Prevalence Trends Over Time in
Denmark: Changes in Prevalence and Age at Diagnosis. Arch. Pediatr. Adolesc. Med.
162, 1150–​1156 (2008).
4. Russell, G. et al. Time Trends in Autism Diagnosis Over 20 Years: A UK Population-​
based Cohort Study.
5. Ibañez, L. V., Grantz, C. J. & Messinger, D. S. The Development of Referential
Communication and Autism Symptomatology in High-​ Risk Infants. Infancy Off.
J. Int. Soc. Infant Stud. 18 (2013).
 Babies and infants 41
6. Teitelbaum, P., Teitelbaum, O., Nye, J., Fryman, J. & Maurer, R. G. Movement
Analysis in Infancy may be Useful for Early Diagnosis of Autism. Proc. Natl Acad. Sci.
U. S. A. 95, 13982–​13987 (1998).
7. Shen, M. D. & Piven, J. Brain and Behavior Development in Autism from Birth
Through Infancy. Dialogues Clin. Neurosci. 19, 325–​333 (2017).
8. Jones, W. & Klin, A. Attention to Eyes is Present but in Decline in 2–​6-​Month-​Old
Infants Later Diagnosed with Autism. Nature 504, 427–​431 (2013).
9. Briggs, H. Autism Detectable ‘in First Months’. BBC News 11 August (2013).
10. van Dijk, W., Faber, M. J., Tanke, M. A. C., Jeurissen, P. P. T. & Westert, G. P.
Medicalisation and Overdiagnosis: What Society Does to Medicine. Int. J. Health
Policy Manag. 5, 619–​622 (2016).
11. Clark, M. L. E., Vinen, Z., Barbaro, J. & Dissanayake, C. School Age Outcomes of
Children Diagnosed Early and Later with Autism Spectrum Disorder. J. Autism Dev.
Disord. 48, 92–​102 (2018).
12. Warren, Z. et al. A Systematic Review of Early Intensive Intervention for Autism
Spectrum Disorders. Pediatrics 127, e1303–​e1311 (2011).
13. Allaby, D. M. & Sharma, D. M. Screening for Autism Spectrum Disorders in Children
Below the Age of 5 years. A Draft Report for the UK National Screening Committee.
Solutions for Public Health. (2011).
14. Chen, I. Understanding Autism: Baby Steps. Spectrum | Autism Research News. www.
spectrumnews.org/​features/​deep-​dive/​what-​baby-​siblings-​can-​teach-​us-​about-​
autism/​(2017).
15. Bumiller, K. The Geneticization of Autism: From New Reproductive Technologies
to the Conception of Genetic Normalcy. Signs J. Women Cult. Soc. 34, 875–​899
(2009).
16. Giovanni, M. A. et al. Health-​care Referrals from Direct-​to-​consumer Genetic Testing.
Genet. Test. Mol. Biomark. 14, 817–​819 (2010).
17. Hedgecoe, A. Schizophrenia and the Narrative of Enlightened Geneticization. Soc.
Stud. Sci. 31, 875–​911 (2001).
18. Latimer, J. The Gene, the Clinic, and the Family: Diagnosing Dysmorphology, Reviving
Medical Dominance (Routledge, 2013).
19. Folstein, S. & Rutter, M. Infantile Autism: A Genetic Study of 21 Twin Pairs. J. Child
Psychol. Psychiatry 18, 297–​321 (1977).
20. Sandin, S. et al. The Familial Risk of Autism. JAMA 311, 1770–​1777 (2014).
21. Brett, D., Warnell, F., McConachie, H. & Parr, J. R. Factors Affecting Age at ASD
Diagnosis in UK: No Evidence that Diagnosis Age has Decreased Between 2004 and
2014. J. Autism Dev. Disord. 46, 1974–​1984 (2016).
22. Daniels, A. M. & Mandell, D. S. Explaining Differences in Age at Autism Spectrum
Disorder Diagnosis: A Critical Review. Autism Int. J. Res. Pract. 18, 583–​597
(2014).
23. Shattuck, P. T. et al. Timing of Identification Among Children with an Autism
Spectrum Disorder: Findings from a Population-​based Surveillance Study. J. Am.
Acad. Child Adolesc. Psychiatry 48, 474–​483 (2009).
24. Sheldrick, R. C., Maye, M. P. & Carter, A. S. Age at First Identification of Autism
Spectrum Disorder: An Analysis of Two US Surveys. J. Am. Acad. Child Adolesc.
Psychiatry 56, 313–​320 (2017).
25. Zwaigenbaum, L. et al. Developmental Functioning and Symptom Severity Influence
Age of Diagnosis in Canadian Preschool Children with Autism. Paediatr. Child Health
24, e57–​e65 (2019).
42 ‘Artefactual’
26. Hosozawa, M. et al. Determinants of an Autism Spectrum Disorder Diagnosis in
Childhood and Adolescence: Evidence from the UK Millennium Cohort Study.
Autism Int. J. Res. Pract. 24, 1557–​1565 (2020) doi:10.1177/​1362361320913671.
27. Sivberg, B. Parents’ Detection of Early Signs in their Children Having an Autistic
Spectrum Disorder. J. Pediatr. Nurs. 18, 433–​439 (2003).
28. Gliga, T., Jones, E. J. H., Bedford, R., Charman, T. & Johnson, M. H. From Early
Markers to Neuro-​developmental Mechanisms of Autism. Dev. Rev. 34, 189–​207
(2014).
29. Landa, R. J. Diagnosis of Autism Spectrum Disorders in the first 3 Years of Life. Nat.
Clin. Pract. Neurol. 4, 138–​147 (2008).
30. Baranek, G. T. Autism During Infancy: A Retrospective Video Analysis of Sensory-​
motor and Social Behaviors at 9–​12 Months of Age. J. Autism Dev. Disord. 29, 213–​
224 (1999).
31. Le Couteur, A. et al. A Broader Phenotype of Autism: The Clinical Spectrum in Twins.
J. Child Psychol. Psychiatry 37, 785–​801 (1996).
32. Mulligan, S. & White, B. P. Sensory and Motor Behaviors of Infant Siblings of
Children with and Without Autism. Am. J. Occup. Ther. Off. Publ. Am. Occup. Ther.
Assoc. 66, 556–​566 (2012).
33. Gallagher, S. & Varga, S. Conceptual Issues in Autism Spectrum Disorders. Curr.
Opin. Psychiatry 28, 127–​132 (2015).
34. LeBarton, E. S. & Iverson, J. M. Fine Motor Skill Predicts Expressive Language in
Infant Siblings of Children with Autism. Dev. Sci. 16, 815–​827 (2013).
35. Sacrey, L.-​A. R., Bennett, J. A. & Zwaigenbaum, L. Early Infant Development and
Intervention for Autism Spectrum Disorder. J. Child Neurol. 30, 1921–​1929 (2015).
36. Leonard, H. C., Elsabbagh, M., Hill, E. L. & Basis Team. Early and Persistent Motor
Delay in Infants at-​ risk of Developing Autism Spectrum Disorder: A Prospective
Study. Eur. J. Dev. Psychol. 11, 18–​35 (2014).
37. Nickel, L. R., Thatcher, A. R., Keller, F., Wozniak, R. H. & Iverson, J. M. Posture
Development in Infants at Heightened vs. Low Risk for Autism Spectrum Disorders.
Infancy Off. J. Int. Soc. Infant Stud. 18, 639–​661 (2013).
38. Watson, L. R. et al. The First Year Inventory: Retrospective Parent Responses to a
Questionnaire Designed to Identify One-​year-​olds at Risk for Autism. J. Autism Dev.
Disord. 37, 49–​61 (2007).
39. Wallace, S., Parr, J. & Herd, A. One in a Hundred. www.autistica.org.uk/​wp-​content/​
uploads/​2014/​10/​One-​in-​a-​Hundred-​Autisticas-​Report.pdf (2012).
40. McGuire, A. E. Buying Time: The S/​pace of Advocacy and the Cultural Production
of Autism. Can. J. Disabil. Stud. 2, 98–​125 (2013).
41. Traustadottir, R. Mothers Who Care. J. Fam. Issues 12, 211–​228 (1991).
42. Eyal, G., Hart, B., Onculer, E., Neta, O. & Rossi, N. The Autism Matrix (Polity,
2010).
43. Green, V. A. et al. Internet Survey of Treatments Used by Parents of Children with
Autism. Res. Dev. Disabil. 27, 70–​84 (2006).
44. Singh, I. Doing Their Jobs: Mothering with Ritalin in a Culture of Mother-​blame. Soc.
Sci. Med. 59, 1193–​1205 (2004).
45. Nadesan, M. Constructing Autism: Unravelling the ‘Truth’ and Understanding the
Social (Routledge, 2005).
46. Armstrong, D. The Rise of Surveillance Medicine. Sociol. Health Illn. 17, 393–​404
(1995).
 Babies and infants 43
47. Armstrong, L. And They Call It Help: The Psychiatric Policing of America’s Children
(Addison-​Wesley, 1993).
48. Beck, U. Risk Society: Towards a New Modernity (Sage, 1992).
49. Foucault, M. Discipline and Punish: The Birth of the Prison (Vintage, 1995).
50. Taussig, M. T. Reification and the Consciousness of the Patient. Soc. Sci. Med. [B]‌14,
3–​13 (1980).
51. Vakirtzi, E. & Bayliss, P. Towards a Foucauldian Methodology in the Study of
Autism: Issues of Archaeology, Genealogy, and Subjectification. J. Philos. Educ. 47,
364–​378 (2013).
52. Green, J. et al. Parent-​mediated Intervention Versus no Intervention for Infants at
High Risk of Autism: A Parallel, Single-​blind, Randomised Trial. Lancet Psychiatry 2,
133–​140 (2015).
53. Turner, L. M. & Stone, W. L. Variability in Outcome for Children with an ASD
Diagnosis at Age 2. J. Child Psychol. Psychiatry 48, 793–​802 (2007).
54. Rutter, M. Autism: Its Recognition, Early Diagnosis, and Service Implications. J. Dev.
Behav. Pediatr. JDBP 27, S54–​S58 (2006).
55. Russell, G., Ford, T., Steer, C. & Golding, J. Identification of Children with the
Same Level of Impairment as Children on the Autistic Spectrum, and Analysis of their
Service Use. J. Child Psychol. Psychiatry 51, 643–​651 (2010).
56. Russell, G. et al. Social and Behavioural Outcomes in Children Diagnosed with Autism
Spectrum Disorders: A Longitudinal Cohort Study. J. Child Psychol. Psychiatry 53,
735–​744 (2012).
57. Lemcke, S., Juul, S., Parner, E. T., Lauritsen, M. B. & Thorsen, P. Early Signs of
Autism in Toddlers: A Follow-​up Study in the Danish National Birth Cohort. J.
Autism Dev. Disord. 43, 2366–​2375 (2013).
58. Shattuck, P. T. et al. Change in Autism Symptoms and Maladaptive Behaviors in
Adolescents and Adults with an Autism Spectrum Disorder. J. Autism Dev. Disord. 37,
1735–​1747 (2007).
59. Whitely, M. et al. Attention Deficit Hyperactivity Disorder Late Birthdate Effect
Common in Both High and Low Prescribing International Jurisdictions: A Systematic
Review. J. Child Psychol. Psychiatry 60, 380–​391 (2019).
60. Fogel, L. S. & Nelson, R. O. The Effects of Special Education Labels on Teachers. J.
Sch. Psychol. 21, 241–​251 (1983).
61. Jussim, L. Self-​Fulfilling Prophecies: A Theoretical and Integrative Review. Psychol.
Rev. 93, 429–​445 (1986).
62. Jussim, L., Palumbo, P., Chatman, C., Madon, S. & Smith, A. Stigma and Self-​
fulfilling Prophecies. Soc. Psychol. Stigma 374–​418 (2000).
63. Rosenthal, R. & Jacobson, L. Teachers’ Expectancies: Determinants of Pupils’ IQ
Gains. Psychol. Rep. 19, 115–​118 (1966).
64. Crais, E. R., Watson, L. R., Baranek, G. T. & Reznick, J. S. Early Identification of
Autism: How Early Can We Go? Semin. Speech Lang. 27, 143–​160 (2006).
65. Reichow, B., Hume, K., Barton, E. E. & Boyd, B. A. Early Intensive Behavioral
Intervention (EIBI) for Young Children with Autism Spectrum Disorders (ASD).
Cochrane Database Syst. Rev. doi:10.1002/​14651858.CD009260.pub3 (2018).
66. Brikell, I. et al. The Contribution of Common Genetic Risk Variants for ADHD
to a General Factor of Childhood Psychopathology. Mol. Psychiatry 1–​13 (2018)
doi:10.1038/​s41380-​018-​0109-​2.
67. Noldus. Aims-​2-​trials. www.noldus.com/​projects/​aims-​2-​trials (2015).
44 ‘Artefactual’
68. Atkinson, A. et al. NIH Biomarkers Definitions Working Group Biomarkers and
Surrogate Endpoints: Preferred Definitions and Conceptual Framework. Clin.
Pharmacol. Ther. 69, 89–​95 (2001).
69. Wolff, J. J. et al. Differences in White Matter Fiber Tract Development Present from
6 to 24 Months in Infants with Autism. Am. J. Psychiatry 169, 589–​600 (2012).
70. Zeman, A. Neurology is Psychiatry –​and Vice Versa. Pract. Neurol. 14, 136–​144
(2014).
71. Folstein, S. E. & Rosen-​Sheidley, B. Genetics of Autism: Complex Aetiology for a
Heterogeneous Disorder. Nat. Rev. Genet. 2, 943–​955 (2001).
72. Persico, A. M. & Napolioni, V. Autism Genetics. Behav. Brain Res. 251, 95–​112
(2013).
73. Yang, M. S. & Gill, M. A Review of Gene Linkage, Association and Expression Studies
in Autism and an Assessment of Convergent Evidence. Int. J. Dev. Neurosci. Off. J. Int.
Soc. Dev. Neurosci. 25, 69–​85 (2007).
74. The Problems with Prenatal Testing for Autism. Spectrum | Autism Research News.
www.spectrumnews.org/​features/​deep-​dive/​the-​problems-​with-​prenatal-​testing-​for-​
autism/​(2019).
75. Nelson, A. Declaration From the Autism Community That They Are a Minority
Group. www.prweb.com/​releases/​2004/​11/​prweb179444.htm (2004).
76. Kapp, S. K. Autistic Community and the Neurodiversity Movement: Stories from the
Frontline (Springer Singapore, 2020).
77. Woman with Down’s Syndrome Takes UK Govt to Court Over Allowing Abortion up
to Birth for Disabilities. Right To Life UK https://​righttolife.org.uk/​news/​woman-​
with-​downs-​syndrome-​takes-​uk-​govt-​to-​court-​over-​allowing-​abortion-​up-​to-​birth-​
for-​disabilities/​ (2020).
78. Dyck, E. & Russell, G. Challenging Psychiatric Classification: Healthy Autistic Diversity
the Neurodiversity Movement. In Mental Health in Historical Perspective: Healthy
Minds in the Twentieth Century (eds. Taylor, S. J. & Brumby, A.) (Palgrave MacMillan,
2020).