The clinical presentation which we call “autism” is now widely believed to be a condition caused my multiple factors with a very variable presentation. In the scientific literature it is frequently called “the autisms”. I would like to present a case which shows that progress in our understanding of autism will depend on our ability to study the very youngest children, that is the newborn, and with the help of newer technologies, eventually the pre-born.
One of my favorite stories, which illustrates the predicament of the research world, is as follows. A drunk came out of the bar staggering around and realized that he had dropped his house key. He saw a streetlight and got on all fours to look for his key. Just at that time his friend was going into the bar, and he stopped and asked him what he was looking for. He told him and his friend wished him luck. Two hours later his friend left the bar and there was our hero still on all fours in the same place. “What are you doing there for two hours?” his friend asked. “Looking for my key” the drunk man replied. “Well it probably is not there if you haven’t found it yet” his friend observed. “Yes I know, but this is the only streetlight around.”
Autism research has been done mostly on older children and adults, because autism cannot be diagnosed rigorously until the age of 2-3 years. So for us the autism diagnosis is the streetlight. The problem is that by that time we can make the diagnosis, the brain has completed so much development, and in a way that is “autistic”. If we could know who will turn out to be autistic much earlier or even in utero, we could intervene with very young children and guide the brain development into a more healthy outcome.
This idea has led many groups to study children at high risk for developing autism. It would be prohibitively expensive to study every child born and see who turns out to be autistic at age 3, but if we can only focus on a subset of children, it is much more realistic. In fact this is what has happened. The most attractive group of high risk children to study is the baby siblings of already diagnosed children. Originally it appeared that there would be a 5-10% rate of autism in the siblings. However, the groups who are doing this work are observing an even higher rate. In addition, many of the children who are not autistic have signs of the “broader phenotype” such as language or social deficits. Much has already been learned about autism through this strategy and there is optimism that much more will come out of these studies.
At the New York State Institute for Basic Research in Developmental Disabilities on Staten Island, we have had a program to study babies who needed to be in the Neonatal Intensive Care Unit. For over a decade we have been following these babies and studying their behavioral and cognitive development. Most of these babies were premature, although some had other problems such as hypoxia or bleeding in the brain. With new techniques available now for saving younger and younger children, we are now facing a new challenge: the developmental problems inherent in the large number of very premature babies who survive. New reports in the past two years coming out of Montreal and Boston have documented that a very large number of very premature babies fail the autism screening tests at 18 months. This does not mean that they will all go on to be diagnosable as autistic, however many of them will. This phenomenon has not been noted in the autism literature primarily because the epidemiologic studies are done on 8 year olds and with the research lag, when we talk about the latest autism numbers it is likely these children were born 10-11 years earlier. It could be that some of the apparent increase in the numbers of autism is coming from children who might have been saved in the neonatal ICU’s.
In our research, we have videos of babies from the time of birth and we have identified 33 of these children who later went on to be diagnosed in the autism spectrum. This is out of over 2,000 babies from the NICU. When we compared those who became autistic to other matched babies we found three predictive signs, which were different in the autism children. These finding were: an inability to visually track an object equally to both sides at one month, an asymmetry of arm and leg strength at one month, and a persistent attraction to high frequency stimulation at four months. If these findings hold up they will represent the earliest predictors that we have of autism or at least the earliest markers we have for high risk children to become autistic.
Although this seems attractive there are still many qualifying factors to consider. One major issue is whether the NICU babies who become autistic are “the same” as other types of autism. In reality this is also true for the baby siblings. There may be many “autisms” and they may not have the same early signs although this remains to be seen. Up until now the baby sibling studies have not looked at children this young and we do not have this type of data.
The other type of early marker that would be very valuable would come from the biology of the child. Biosamples such as the placenta, cord blood, the mother’s blood, etc. might yield markers. However, the studies to look at these samples and correlate them with the development of the child have not been done. Our goal in NY State is to focus on “birth cohorts” and look for the early signs of autism or other developmental disabilities. We are hoping to set up many sites and study the issues around pregnancy (such as medical risk, stress, etc.), the environmental influences, and the signs which might be observable in early infancy. In fact there are more types of high risk children and capturing all of them would give us a much better handle on what causes autism as well as the natural course of the disease.
The reason for doing all of this, ultimately, is to reduce the burden of the disease and to improve the lives of those affected children and their families as much as possible. In order to accomplish this there would need to be a treatment component along with these studies. Instituting very early intervention would be of the utmost importance and would have the greatest possibility of having a significant effect on the outcome for a given child. The first two years is one of dramatic development of the brain and it is possible that, through various means (educational, environmental and pharmacologic), we can influence the outcome.
This leads to another scientific question which is in need of answer. How much of the “autism” is present before birth and how much of it occurs in the first two years of life? This is not an easy question to study. We do know from brain tissue research that many of the brain changes noted in autism are findings which could only be explained by in utero changes, some of which occurring as early as the first trimester. This does not however rule out that there are further developmental changes which continue along with the brain development of the child. For example, suppose a child brain at the age of 2-3 months has a problem which prevents that child from processing or modulating sensory information. The child may then “learn” not to focus on speech sounds. It may be that there was nothing inherently wrong with the language parts of the brain, but due to the other problems, the language centers never fully kick in and do their function. Child development studies for many decades have documented that various problems with nurturing such as the Romanian orphans during the Ceausescu regime developed a syndrome very similar to autism. It is likely that these children do not share the biologic underpinning of the autism cases we have here in America and so there is something about severe deprivation early in life that leads to autism symptoms. A question to be posed is whether a biologic lesion is causing a self imposed isolation in the child leading to the syndrome which we call autism. This and other similar questions can begin to be investigated if we move the age down at which we start studying autism. With the support of the advocacy community, the state agencies, and the voluntary sector there is no reason why these goals cannot be achieved.