Autism prodrome 35 of 89
emphasizes the notion that whilst regression may be part of the prodrome and the
clinical picture for children with ASDs it will not be a universal one. One note of
caution, however, is that although there is at least one example from the prospective
high risk sibling studies where a subgroup of children who went on to have ASDs lost
skills in the second year of life (Landa et al., 2007), most studies have not yet
provided prospective data on the regression phenomenon (though see Ozonoff et al.,
in press; for an exception where loss of skills was common in a group with an ASD
outcome). It also appears that objective observations and parental reporting of
regression or setback do not always concur (see Ozonoff et al., 2008, in press; for
discussion).
However, regression remains intriguing as a candidate prodromal feature of
ASD because, aside from later occurring childhood disintegration disorder (CDD)
where regression occurs after the age of 2 years or even later following clearly typical
development (regression in autism is before this age commonly between 12 and 20
months of age) (Volkmar & Cohen, 1989) or Landau-Kleffner syndrome that is due to
localized peri-sylvian epilepsy (Robinson, Baird, Robinson, & Simonoff, 2001), it
appears to be relatively specific to more narrowly defined autism. A recent report
examined how common regression was in children with specific language impairment
(SLI), also called developmental language disorder (DLD). Pickles et al. (2009) found
that regression occurred in only 1% of children with SLI compared to 15% of children
with autism or a broader ASD. In a related study the same group reported that
regression occurred in 30% of children with narrowly defined autism (children
meeting ICD-10 criteria for childhood autism plus the autism threshold on the ADI-R
and ADOS), 8% of children with broader ASD/PDD and only 3% of controls with a
mixture of intellectual, learning and behavioral difficulties (Baird et al., 2008).