What Are the Probable Earliest Signs of Autism?

Of the comments and queries I receive on this blog, the greatest number are from parents concerned that their babies are showing signs of autism. Of questions about autism, almost all of them focus on a lack of eye contact—the babies do not gaze at their parents’ faces as much as the parents expect them to.

Most parents, and certainly many Internet authors who discuss autism, assume that whatever are the signs and symptoms of autism in older children and adults, those will also be the signs and symptoms of autism in young infants. They know that social awkwardness and a lack of eye contact and other communicative gestures are common among older children with autism spectrum disorder (ASD), so they assume that infants who are fated for an ASD diagnosis will also lack eye contact. But in making this assumption they miss two important points.

The first point is that infants in the first two or three months are not easily attracted to pay attention to people. They will do it now and then, but often they respond only to quite dramatic adult facial expressions with wide open-mouthed smiles. Years ago, this developmental period was referred to as a stage of normal autism—the word “autism” deriving from the Greek word for “self”, and the babies being focused on themselves rather than the environment. Now that people are terrified about ASD, one doesn’t come across this expression, normal autism, any more, but that doesn’t mean that the stage no longer exists. What would be a symptom of ASD in an older child is a sign of perfectly normal development in a young infant. There is no point in expecting a baby of a few months to make extensive eye contact, any more than there would be any point in expecting her to build a tower of two blocks or to spoon-feed herself.

My second point is that earlier and later behavior patterns may be remarkably different in cases where there are developmental problems. A good example is the pattern shown in Williams syndrome, a genetic syndrome resulting from loss of certain parts of a chromosome. Williams syndrome is not terribly debilitating, but it does cause developmental changes that are rather different from typical development. Young babies with Williams syndrome are terribly colicky and cry frantically no matter what is done to soothe them. They are not interested in other people at that point. But when the colicky stage passes, they become extremely interested in people, stare at them intently, and appear to be “starved” for eye contact. We might expect them at this point to be very interested in communication and to speak early, but no; in fact, their speech is delayed by about a year. Once they do start to speak, they became chatty conversationalists. As adults, they are still talkative, with wonderful language abilities, and highly sociable—but socially awkward at the same time because they seem to lack the social anxiety that guides most of us. The screaming colic and delayed speech of the younger Williams syndrome individual are by no means symptoms of either the strengths or the weakness of the Williams adult—and it’s very possible that a similar situation holds for ASD people, whose later symptoms may not mirror the earlier ones (if there are any).

A possible conclusion from these two points is that the current preoccupation of parents with eye contact may be irrelevant to the diagnosis of autism. The fact that older ASD children may not use eye contact for communication very much does not mean that we can identify infants who will later be diagnosed with ASD by looking for them to make eye contact.

A recent paper on autism  provides some interesting insights into possible early symptoms of autism. (Thomas, M., Davis, R., Karmiloff-Smith, A., Knowland, V., & Charman, T. (2016). The over-pruning hypothesis of autism. Developmental Science, 19, 284-305.) This is a very complicated paper, and I am only going to refer to one of its points here.

The Thomas paper is one that discusses an idea about how ASD develops. The basic idea is that autism results when a particular problem occurs during early development. It is well known that during the first year, there is great overdevelopment of synapses or connections between neurons in the brain, followed by disappearance of many that are little used—a process sometimes called “pruning”. Although some authors have suggested that autism results from too little “pruning” of synapses, Thomas and his co-authors hypothesize that too much “pruning” could be the problem. They have tested this hypothesis by developing a neurocomputational model to allow them to predict what kinds of problems should result from excessive “pruning”.

Like all good scientists, Thomas and his colleagues are testing their model against some longitudinal studies of development of autistic children, and to some extent are finding that the longitudinal studies show the symptoms they predict on the basis of the timing of pruning events during early development. These do NOT include symptoms of social interaction problems. They do include difficulties with sensory development like over- or under-sensitivity to sound or touch stimulation, and difficulties with motor development. Motor development problems as a precursor to autism have been discussed since the 1980s, when studies of home videos were sought by researchers as a way to see the early development of children later diagnosed with ASD. Even before that, clinicians had noted that unusual movement patterns like crawling asymmetrically or always reaching with one hand during early childhood were related to a variety of later developmental problems. These sensory and motor problems in the first year or two may indicate that children will later show the social interaction problems often associated with autism—even though the children when younger do not show unusual social interactions.

The sensory and motor forerunners of autism are not yet clearly understood, so they cannot be used for accurate  identification of “pre-autistic” babies. In addition, many young children who are thought to be autistic at age 2 show normal development later. Much as we might like to have early identification and early intervention, we don’t have it yet. But if identification and intervention are ever going to work, they will have to be focused on development that is really not typical—and it is quite typical for babies in the first weeks and even months to look at things other than faces a good deal of the time.