Bengt Sivberg

The efforts of our ongoing studies are to describe observed early atypical behaviours

during infancy, especially those that may be associated with autism spectrum conditions

(ASCs) in order to make it possible to start an early support intervention.

During the last decade a number of studies about early screening of ASCs have been

published. Mainly they have used two types of sampling, either a general screening

tool followed by ASC-specific instruments/observations/interviews for children

scoring positive at the first line, or alternatively, an ASC-specific tool for children in

the risk zone (sibling or twin studies) (Pinto-Martin et al., 2008). The number of

studies is quantifiable and also includes studies using a randomised controlled experimental

design (Dawson et al., 2010). However, screening studies on the basis of a

given population (Nygren et al., 2012) in primary health care or at paediatric clinics/

university institutions are sparsely reported in scientific publications leaving us

without reliable knowledge of ASC prevalence in very young authentic populations.

Recently Bolton, Golding, Emond, and Steer (2012) published a longitudinal and

cross-sectional study of early predictors of later ASCs and autistic traits. Results

indicated that in the first year differences in attention, inhibition of control and vocalisations

were striking in comparison with typically developed infants (Landa, Gross,

Stuart, & Faherty, 2013).

The American Academy of Pediatrics, in its Embargo of 2006, claims that early

identification of children at risk of ASCs is urgently needed for early intervention.

The reasoning behind that statement is the relative plasticity of the developing brain

during a child’s first years and also the increasing amount of knowledge of early

signs and markers of ASC-specific behaviours (American Psychiatric Association,

2013; Barbaro & Dissanayake, 2013; Barbaro, Ridgway, & Dissanayake, 2011;

Crane & Winsler, 2008; Daniels, Halladay, Shih, Elder, & Dawson, 2014; Dereu

et al., 2010; Gura, Champagne, & Blood-Siegfried, 2011), atypical child development

(Ozonoff et al., 2008) and medical status associated with ASCs (Eriksson et al., 2013;

Karmel et al., 2010). The problem of regression (Ozonoff et al., 2010; Toth, Munson,

Meltzoff, & Dawson, 2006) during the second year of life for about 20% of children

later diagnosed with an ASC has been highlighted as a contra argument (Webb,

Dawson, Bernier, & Panagiotides, 2006) towards a consequence of too-early screening.

This argument is relevant for national child health-care programmes that follow infants

only few times during childhood, but if one has a continuous programme up until

school age, these children would show up in an ASC programme later anyway.

Although the risk of false-positive findings exists, most parents prefer the stress and

worry of an early positive identification in contrast to a missed opportunity of an

early intervention (Barton, Dumont-Mathieu, & Fein, 2012). However, it is a

problem to distinguish children with other developmental concerns from children

with ASCs at around 12 months of age (Manning-Courtney et al., 2013).

Inglese (2009) emphasised the need for nurses in the community and primary care

settings to improve their knowledge about ASCs and to learn to screen for behaviours

associated with ASCs in children due to the fact that they will encounter an increasing

number of children with these conditions. Cangialose and Allen (2014) emphasise the

support from recent research that ASC-specific behaviours and developmental delays

emerge early in infancy and that an early screening can contribute the identification

of infants in the risk zone of an ASC. Eriksson et al. (2013) recently reported from a

Swedish intervention programme where toddlers (20–54 months) received intervention

care based on principles of non-intensive or intensive applied behaviour analysis. The

result indicated that toddlers without epilepsy and intellectual disorder (ID) gained

more positive effects of the intervention in comparison with children with epilepsy

and ID. The Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition

(DSM-5; American Psychiatric Association, 2013) is the most widely accepted nomenclature

used by clinicians and researchers for the classification of mental disorders

manual focuses on two diagnostic domains: social communication deficits and

restricted and stereotyped behaviours. Deficits in social skills and delays in language

development are probably the two most reliable aspects of very early identification

of ASCs, but a social communication delay must not obligatorily refer to an ASC diagnosis

and is not sufficient for ASC identification. Moreover, there is also a need for

identification of restricted and/or stereotyped behaviours. It may be difficult to find

reliable behaviours of these ranges during infancy.

An enlightening perspective among both clinicians and researchers is that ASCs

often are associated with other medical problems (Carbone, 2013; Konopka, 2013).

Teitelbaum, Teitelbaum, Nye, Fryman, and Maurer (1998) used a movement analysis

system to compare movements (lying, righting oneself, sitting, crawling and standing)

of infants with an ASC with typically developed infants at the ages of four and six

months. They concluded that movement disturbances are clearly detectable at this

early age. Ozonoff et al. (2010) found indications of an overall slowed rate of motor

development for infants with an ASC compared with typically developed infants. Seizures

have also been observed during the first few days of life and indicate a more

serious development of an ASC (Eriksson et al., 2013). At one month of age a

higher percentage of infants later diagnosed with an ASC displayed asymmetric

visual tracking and abnormal upper extremity tone, and at four months they showed

a greater preference for higher frequency stimulation when less aroused as well as performing

more like newborns than four-month-old infants (Karmel et al., 2010). At six

months of age, Chawarska, Macari, and Shic (2013) found that infants later diagnosed

with an ASC paid less attention to the social scene and were more diminished in their

ability to react spontaneously to people than typically developed infants. Atypical patterns

of electroencephalography activity have been demonstrated by Dawson et al.

(2002) in young children with an ASC. They have shown that early intervention

with the Early-Start Denver Model in a randomised, controlled trial design study

resulted in increased cortical activation in the ASC group of children that viewed

faces in comparison with children with an ASC group treated as usual. Research has

also been directed at infants’ attention capacity in order to study both audio-visual

and visual stimulation and its effects on memory (Flom & Bahrick, 2010) as this is

essential for typical psychological development of infants.

Screening for ASCs in primary health-care settings was performed by Robins as

early as 2008 by using the Modified Checklist for Autism in Toddlers (M-Chat) and

following the M-Chat with an interview (Robins, 2008). The M-Chat instrument is a

further development of the original British instrument the CHAT (Baron-Cohen,

Allen, & Gillberg, 1992). In recent publications, Robins et al. (2014) and Chlebowski,

Robins, Barton, and Fein (2013) have advanced the M-Chat R into R/F for toddlers

between the ages of 16 and 30 months in a large sample (n = 16,071; 18,989, respectively)

of parents/toddlers in the American city of Atlanta and state Connecticut. Results

indicated that 47.5% of the children having 3 or higher CHAT points were in the risk

zone of being diagnosed with an ASC. A recent Swedish study (Nygren et al., 2012)

completed with the M-Chat and the joint attention observation Five-Item Observational

Scale with an additional brief speech/language assessment scale have been performed

in primary health-care settings with children of the ages 30–34 months. Nurses

observed children’s competencies of joint attention at 2.5 years of age. Sixty-four children

(of 3999) were purportedly autistic and 48 were confirmed with an ASC diagnosis.

Other studies have been performed with infants and toddlers in primary health-care settings

or community-based settings. Barbaro et al. (2011) conducted a screening study in

Australia using paediatric nurses in the first level of implementation of the developmental

surveillance programme with the purpose of lowering the age at which ASCs are

diagnosed in order to start an intervention an earlier age. The conclusion of the

study was that routine monitoring done by nurses for ASCs should be standardised

in relation to what it is now. Barbaro and Dissanayake (2013) reported a new community-

based study from Australia with the focus of determining the most discriminating

and predictive markers of ASC. The Social Attention and Communication Scale was

used at 12, 18 and 24 months of age for three groups: children with autistic disorder,

ASC and developmental and/or language delay. Deficits in eye contact with finger

pointing were consistent across the ages for the groups of autistic disorder and

ASCs, however at the age of 24 months pretend play was the best marker.

In a study from the Netherlands (Oosterling et al., 2010) using the Early Screening

of Autistic Traits Questionnaire the aims were to develop and evaluate a two-step

screening approach for early detection of ASCs. Results showed that in the experimental

study group ASCs were diagnosed 21 months earlier than in the control group and

with the mean age at ASC diagnosis decreasing by 19.5 months. Children from the

experimental group were 9.4 times more likely to be diagnosed before 36 months

than the controls. Pinto-Martin et al. (2008) conducted a study comparing the

number of children (18–30-month-olds) identified at risk for ASCs visiting primary

paediatric care using a general screening tool (Parents’ Evaluation of Developmental

Status) and an autism-specific tool (M-Chat). Findings indicated that a parent-based

tool was not sufficient in comparison with an autism-specific tool. However, the

sample was too small for any far-reaching conclusions to be made.