Infants, Autism,and "Eye Contact"

I notice on my blog information this morning an unusually large number of reads of an old post, http://childmyths.blogspot.com/2011/07/eye-contact-with-babies-what-when-why.html. I would guess that this interest is because of various announcements of a research article in Nature by Warren Jones and Ami Klin, “Attention to eyes is present but in decline in 2-6-month-old infants later diagnosed with autism” (on line, 6 November 2013). A piece on the front page of the New York Times discussed the Nature paper (http://well.blog.nytimes.com/2013/11/06/a-babys-gaze-may-signal-autism-study-finds/).

I hope everyone will understand that this research does not make it possible for parents to assess their young baby’s typical or atypical development by seeing whether they “make eye contact”, that is, engage in mutual gazing for short periods with another person, or visually track another person’s eyes. This study, valuable as it is, does not give guidelines for evaluating a baby’s gaze in an everyday social setting.  All babies, like all older persons, look at eyes sometimes, at mouths sometimes, at other parts of the face sometimes—and away from a person at other times. Whether they do “mutual gazing” with another person depends to some extent on what the other person looks like or does, whether the distance and illumination are favorable, whether the baby is tired, etc. An important point made by Jones and Warren is that differences in eye-tracking are not present at birth, when babies do not reliably look at eyes, but are seen after two months, when typical development involves an increase in attention to eyes so that by 6 months typically-developing babies look at eyes as much as adults do.  

In any case, the babies in the Jones and Klin study  did not “make eye contact”; there was none to be made, because the babies were watching a video of another person, not engaging face-to-face with a real person who could look back at them. The question was whether recordings of their eye movements showed that the babies looked at the videorecorded eyes or at other parts of the face or the screen, and how long their gaze lingered on a particular area. Typically-developing babies begin to look at the eyes more and more as they mature from about 2 months to about 6 months, but it appears that for babies who will later be diagnosed as autistic, this developmental process is “derailed” during that period.

The Times report quoted Warren Jones as saying that eye-tracking (following another’s eyes with one’s gaze) is only one channel that may be related to autism. He proposed that social communication through  touch or listening is probably similarly related to autism, but these processes are more difficult to measure than eye movements. (This suggests, by the way, that a treatment that increased eye contact might not necessarily affect autistic development, because the atypical gaze may be only one symptom of a central problem.)   

Naturally, the concern with early diagnosis of autism is associated with the hope that early intervention would get the “derailed” social development back on track. The New York Times report included the suggestion that intensified or pleasurable social experiences might be helpful if they were provided during the possible sensitive period or window of opportunity that may be indicated by the Jones and Klin study. Does this suggestion imply that parents of autistic children have done something wrong, or failed to do something right? Does the strong genetic factor at work in autism determine that these infants need unusual care given by their parents, different care than is needed by more typically-developing babies? How do the interactions between genes and environment guide development?

There are several ways in which genes and environment can interact. One of these, a passive interaction, means that the environment directly affects development in different ways depending on a person’s genetic make-up. For example, if a genetic problem made a baby not very sensitive to social signals like gaze, and if those signals needed to be experienced for good development, more intense or noticeable signals might help development. This is an idea that has been around for some time. Nikolaas Tinbergen, the Nobel Laureate in Medicine, suggested that people working  with autistic children should wear masks with big eyes to get the children’s attention. Some alternative treatments for autism have parents wear large eyeglasses and hold objects beside their eyes when they offer the objects to autistic children. There is no evidence that this is effective, though, so if there is a gene-environment interaction in autism it may not be this kind, and treatments that use these methods will probably not be helpful no matter how early they are provided.

Gene-environment interactions may also be evocative. This means that the genetic material causes the individual to act in ways that make some experiences more likely to happen. This might mean that when typically-developing babies pay attention to eyes, they attract adults to socialize with them, and that socialization facilitates good development. If babies who will later be diagnosed as autistic do not pay attention in that way, or do other things that attract adult play and interaction, they might not receive social interactions that are needed to facilitate development. In that case, treatment might involve helping parents become more responsive to any cues the babies give, or more able to find ways of interacting even though the baby doesn’t “give the signal”. This might or might not need to happen during a particular early period, although the Jones and Klin study seems to suggest that there are important changes early on.

Gene-environment interactions can also be active. This means that the genetic material causes the baby to do things that naturally provide certain experiences--  like moving in ways that strengthen muscles and increase flexibility, or preferring some foods over others. Children who find dancing or singing or talking easy and fun will do those things over and over and get even better at them as they actively seek these experiences. Toddlers with Williams syndrome are said to seem to have a “hunger” for eye contact, and although they are initially slow to talk they later become extremely sociable and communicative. If this were the case in autism, it may be that typically-developing children are genetically determined to do things like looking and imitating that encourage good development, and autistic children lack that determinant. In that case, it would be a matter of a great deal of work to determine what it was that babies typically did that gave them the “right experiences”, and to figure out how such experiences could be given to autistic children. Even then, it would have to be shown empirically that the treatment was effective.

Finally, it’s possible that there is little or no gene-environment interaction. The Jones and Klin finding may simply show a developmental trajectory that is different for typical and autistic children because of their genetic differences. The increasing differences between 2 and 6 months might, unfortunately, not indicate a sensitive period, but just result from innate developmental patterns. In that case, early diagnosis might or might not help interventions work.

There is a lot of  other interesting work that Jones, Klin, and others have published, and I hope to be able to comment on it soon.