Dr. Ami Klin is the Chief of Autism and Related Disorders of the Marcus Autism Center. Dr. Klin is also a Georgia Research Alliance Eminent Scholar at Emory University and director of the Division of Autism and Related Developmental Disabilities in the Department of Pediatrics at Emory University School of Medicine. Dr. Klin is also on the Autism Science Foundation’s Scientific Advisory Board. ASF Intern Max Rolison interviewed Dr. Klin about his research on high-risk siblings and eye-tracking.
Max Rolison: What got you interested in autism?
Ami Klin: I have worked with children with disabilities all my adult life. I was always involved in clinical care. During my graduate years I lived in a residential unit for adults with autism in London. I was very invested personally in the clinical care for individuals with autism. From an academic standpoint, I was always interested in the science of social development. I saw a convergence between those two fields. So I basically married the two and became a clinician and hopefully a scientist in the areas of social relationships and social development.
MR: How did you get involved in studying high risk siblings?
AK: For many years, we didn’t see any children who were very young because children would come to a place like Yale only at a certain age. With much greater awareness of autism, thanks in part to the work that Alison Singer has done for so many years, parents began to reach out to us with younger and younger children. Autism is a neurodevelopmental disorder that that should be diagnosable in the first years of life. And yet, all of the knowledge we had of those first years of life were based on the retrospective information that one would get from either parents reporting or some indirect source like the videos that were made of children prior to the time that the children were diagnosed. Of course, in order to understand autism, it is absolutely critical that we understand the first months and years of life. Although autism has a strong genetic basis, the causes are numerous and varied.
We still need to answer the question of why we have so many different causes and yet we still have a unitary symptom in the sense that a clinician like myself can walk into a playroom and recognize a child with autism as different from children with other developmental disabilities. In other words, we need to ask where does the homogeneity of autism come from. And there is no replacement to basically accompanying children from the time they are born. Even though the prevalence of autism is very high, say 1 in 110/120, it would not be possible for one to create a risk based sample on the basis of the general population because you would have to see too many children in order to see one child with autism. We know that the recurrent rate among siblings is much higher than in the general population. This is why it is imperative to follow younger siblings, and, in fact, we now know that this recurrence is even higher than we expected in the beginning. It is important for science – unveiling the mystery of the first 2 years of life – but it is also important to families, who are typically very concerned about their children once they have a child with autism. What our program did that others did not is that we decided that we were going to follow children in the first year of life very intensively. Most other centers see siblings only once in the first year of life. We decided we were going to follow the children monthly for the first six months, and every three months thereafter. And that was a critical decision, making possible the results we eventually got.
MR: Do you continue to follow siblings after they are either diagnosed or not diagnosed?
AK: By all means. Our model has always been, both at Yale and here in Atlanta at the Marcus Autism Center, to strike a very strong relationship with the families from the time that the child is very small, and we are there for the duration. And that means we retain a relationship with those children throughout their life. What we attempt to do is to accompany those children long after the time that they either receive or do not receive a diagnosis. If we take this model outside of the context of the baby siblings project, (because that is a very recent development, something we only started 5/6 years ago,) during my tenure at Yale – I was there for 21 years – many of the toddlers that I saw are now young adults. Nothing really replaces that kind of lifetime perspective in order for one to understand the origins of autism and the outcomes of autism. In fact, one of the reasons I came to Atlanta, going back to the time that I lived in that residential unit for adults with autism many years ago, those were individuals who spent a great deal of time in an institutional environment. Those were individuals who were extremely disabled, and those were individuals who had never been treated. At that time I thought that autism was a very intractable condition.
More recently, by following children from birth, we realized the power of neuroplasticity, or the maleability of the brain in the first years of life. And that is why I feel so invested in infancy work – in fact I believe that it is a bioethical imperative to identify children early and intervene early, as there is a great opportunity to optimize their outcome. In fact, I’ve never felt as optimistic in my life. My feeling is that we can significantly attenuate the condition by identifying them early and intervening early.
MR: What about the high risk siblings who do not receive a diagnosis?
AK: When you think about high-risk siblings, their outcome usually falls within one of four categories. One, they also develop autism. Two, they develop some signs of autism, without necessarily meeting criteria for the condition. Three, they may have some developmental difficulties, but not something that is central to autism. Four, they are unaffected siblings. It’s a little bit early in this line of research for us to make definitive statements about each one of those categories. Right now, our findings are that those who eventually develop autism are definitely different from their typical peers, but there are some signs that unaffected siblings may have some things that remind us of the genetic liability they share with affected siblings Some colleagues of mine, Kevin Pelphrey and Martha Kaiser, conducted a very interesting fMRI study in which they found the performance of unaffected siblings was more aligned with individuals with autism than with controls. Collectively there is a very optimistic message here: were we to find out why, despite a shared risk, some siblings develop autism while others don’t, then we might be able to diminish risk and augment resilience, thus possibly significantly attenuating the disability associated with autism in the affected children.
MR: What can we learn from eye tracking that we may not learn from other methods?
AK: Well we don’t have a very good way of quantifying the social disability in autism in a way that is objective. What eye-tracking provides us with is the quantification of very early emerging skills that are critical for socialization, and it’s a very noninvasive and low cost procedure. We can obtain highly quantified, moment by moment, measures of engagement of children with the social world, and deviations thereof become red flags for autism. This is not quite possible for the vast majority of other cognitive science methods available today. We have excellent group findings in EEG, we have excellent group findings in fMRI, and we even have excellent group findings in structural brain imaging. But those measures tend to be group measures, not something that can allow us to eventually develop a procedure that is both diagnostic and prescriptive in autism. And yet, there has been some progress recently in this regard. What we’ve been learning from the use of eye tracking methods, and the development of novel data analytic strategies is that we can make quantified statements about a child’s behavioral diagnostic classification, as well as profile of social learning in a way that appears to have meaning for the individual child. In other words: these measures appear to be diagnostic and prescriptive.
MR: Can you discuss the basis behind the study presented at IMFAR?
AK: That study captures the very first wave of analysis that we’ve conducted on our baby siblings data using eye-tracking methods. About 110 children were followed from the age of 2 months through the age of 36 months. There were 12 data points, 5 data collection points before the age of 6 months. And the construct that we use is very familiar to pediatricians: ‘ growth charts’. If you take a child to a pediatrician, the pediatrician will measure weight and height. If there is a significant deviation from the norms, there is reason for concern and reason for a follow-up. What we have been trying to develop is a growth chart of social development through quantified, densely sampled measurements of social visual engagement. The study showed that children deviating significantly from normative growth charts at the age of 6 months were much more likely to be diagnosed with autism at 36 months.
MR: What are you currently researching?
AK: We are researching a whole range of things. In the realm of infancy and toddlerhood, our major goal is to create a system of measures that are independently validated markers of autism, which, together, converge into a robust profile of risk for the condition. The model is medicine in general. If somebody is concerned about a child’s disease, you may take a blood sample that is sent to a laboratory, and one sample provides results on a range of, say, proteins, and it’s the profile of those various proteins that will tell you whether the child has or does not have a given disease. A major thrust is to have an inexpensive behavioral assay that is easily obtained, in our case experiments using eye tracking technology, and then using those data to derive a series of independently validated but convergent measures of risk for autism.
As I mentioned, we want these measures to have individual significant – being both diagnostic and prescriptive – they should inform treatment. Those measures involve both the visual environment—how children visually engage with the world around them—and how they engage with the world vocally. Eventually what we need is highly quantified growth charts of social engagement or social development upon which we can map biological measures that can be obtained in the first year of life. The major thrust of our work is to try to bridge the gap that exists between genes and behavior/symptoms. And that gap, in order to be bridged, needs to be filled with much more advanced knowledge of social development and neuroscience in the infancy period.
MR: What do you think needs to be researched more to better understand autism?
AK: I think that we need to understand more how children with autism build their understanding of the world. Most research that you see published focuses on deficits exhibited by children with autism: in other words, how the system breaks down. In my view, it is critical that we also understand how the system builds up. We know that children with autism learn about the world differently than typical children. What we don’t know as well is how they go about doing so. We have many preconceptions about the way their learning takes place, but I think we need to have a much better understanding of how this happens in actuality, and from birth on because the development in the first years of life is really critical shaping, in many ways, subsequent learning, compensatory but also maladaptive behaviors. In a way, I feel that we need to objectify what we all do clinically. As a clinician, I often feel that I am a walking laboratory of social engagement. We need scientific techniques that quantify/objectify that intuitive resonance that we all use instinctively as clinicians. So many aspects of social adaptation are still so poorly understood. And this is despite of the fact that, evolutionarily, these skills are foundational for survival, are highly conserved – we see them in many species other than man – and they are developmentally very early emerging – babies display them in the first weeks and months of life. These skills, which were once important for survival, now set social interaction in motion, which, in turn, becomes the platform for the development of social mind and brain. Individuals with autism can reach incredible heights with their skills in some areas; and yet, they universally have difficulty with these foundational skills. Were we to fully understand this gap, I believe, we would be much better able to support their growth in social relationships, communication, vocation and independence.
MR: What steps have you taken to build up the Marcus Autism Center?
AK: The Marcus Autism Center is the largest center for clinical care for children with autism and their families in the country – 4,825 children seen in 2010, with vast treatment centers. It is now affiliated to Children’s Healthcare of Atlanta, the largest system of pediatric healthcare in the country, and the Division of Autism & Related Disorders at Emory University School of Medicine. Founded by Bernie Marcus in 1991, it is now aspiring to become a hub of clinical science, with innovative labs focused on social development, speech science and social neuroscience, a large standardized clinical assessment service making possible large treatment trials – from behavioral to psychopharmacological modalities, and a web of collaborations with Emory Human Genetics, Yerkes National Primate Research Center, the CDC, Georgia Tech, among others, which push the limits of our current knowledge by investigating gene-behavior, gene-brain and brain-behavior relationships associated with autism and at the heart of social neuroscience, with a very strong translational emphasis. We will measure our success not by the number of children entering our doors but by the number of families impacted in the community. Hence our various partnerships with community stakeholders, and our focus on projects that bring science to pediatricians’ offices and early intervention providers. Empowered by Children’s Healthcare of Atlanta – science with an entrepreneurial spirit, we are now working on enabling primary care physicians and providers with early detection, case-management, family support and treatment tools and training. Our goal at Marcus, Children’s and Emory is not only to generate new science – we aim at creating a new system of healthcare delivery for young children with autism and their families. This is a model that will hopefully create better outcomes by the time the children enter school eligibility years. These activities are all organized as research projects as our goal is to create replicable models that can be adopted in primary care settings anywhere, beginning with first steps toward universal screening for autism in infancy. It’s enormously exciting, and our rate of progress is a little dizzying. Personally, it is wonderful to be actually doing things that we always dreamt of doing but could not muster the institutional and community commitments or the necessary resources.