DSM-IV Criteria and Early Diagnosis
Children with autism and ASD over the age of 3 generally display characteristics of autism that map well onto the diagnostic criteria outlined in the Diagnostic and Statistical Manual of Mental Disorder (DSM-IV) (2000). It is not clear that DSM-IV criteria are equally applicable to children under age 3 and particularly younger than 2, however. Stone and colleagues (1999) have examined this question empirically in a group of two-year-olds diagnosed clinically with autism.
To address concerns about the difficulties with the DSM-IV for the early childhood years, the Diagnostic Classification of Mental Health and Developmental disorder of Infancy and Early Childhood: Revised Edition (DC:0-3R) (2005) was developed and published by the Zero to Three organization. It is an evolving classification system that is based on expert consensus rather than empirical findings and is designed to complement, not replace the DSM-IV in identifying early mental health and developmental difficulties in children under the age of five. The classification system adds additional diagnostic categories not found in DSM-IV, for the purpose of clinical guidance and more detailed description of early childhood difficulties. Since it has not been empirically validated, it should not be substituted for the DSM-IV as a diagnostic standard for ASD.
A great deal of research has been done to explore the earliest signs of autism and when they can be detected. Several research strategies have been used to study these questions, including:
Strong consensus has been obtained across all research methods. Behaviors that consistently discriminate infants or toddlers with autism from those with non-autism developmental delays or typical development are orienting to name, eye contact, social referencing, interest in other children, joint attention, affect sharing, and imitation (see Rogers, 2009 for a review).
A 2010 study (Ozonoff et al.) that compared siblings of children diagnosed with ASD and typically developing infants found that the earliest characteristics of autism are not readily observed at six months, but emerge gradually between nine and twelve months of age. Observed characteristics, such as a lack of shared eye contact, smiling and communicative babbling, contribute to the consensus reached by previous research.
Recently, our understanding of how the first signs of autism emerge has changed. The traditional notion was that there were two broad onset patterns.
The most common course involves early signs of developmental differences, before 12-18 months of age. In approximately a third of cases, however, symptom onset does not occur until sometime in the second year of life, between 14 and 24 months of age (Fombonne, & Chakrabarti, 2001; Lord, Shulman, & DiLavore, 2004), after a developmental regression, or loss of previously acquired skills. The most frequently reported skill lost is language, although virtually all children who lose language lose social behaviors as well, such as eye contact, social interest, and engagement with others (Ozonoff, Williams, & Landa, 2005). However, more recently, we have come to realize that these two onset patterns do not capture all of the different ways autism emerges.
There seem to be different patterns of late onset of symptoms - regression and plateau - though it is often difficult to separate children who show a regression (skill loss) from children who show a developmental “plateau,” marked by a failure to progress and transform simpler behaviors into more advanced social and communication skills (Chawarska et al., 2007; Siperstein, & Volkmar, 2004). Moreover, many children with a reported regression already have a number of developmental delays prior to the regression itself (Goldberg, Thorsen, Osann, & Spence, 2007; Ozonoff et al., 2005; Siperstein, & Volkmar, 2004; Werner, Dawson, Munson, & Osterling, 2005), and some children with clear signs of autism in the first year of life also experience skill loss in the second year (Werner, & Dawson, 2005).
Some have suggested four different onset patterns: early (before 12 months) onset of symptoms without skill loss, late onset of symptoms seen in skill loss (regression) without early symptoms, early onset combined with later skill loss, and late onset without skill loss (developmental plateau) (Ozonoff et al., 2010).
Another way to think about how autism begins is that children may reach the threshold for diagnosis at different points in the first three years of life, involving different amounts and/or combinations of early signs, slowing development, lack of progression, and frank losses (Landa et al., 2007; Rogers, 2009; Ozonoff et al., 2010).
View two videos that illustrate the two traditionally defined patterns of onset:
Remember that not all children will fall into these categories or that failure to display these kinds of trajectories does not rule out an autism diagnosis.