– Okay, so thank you very
much for coming today. And welcome to the
introductory lecture on autism. In this lecture, we’ll focus
particularly on how autism expresses itself in the
first three years of life. We’re going to talk about core features, comorbid characteristics. We’ll talk a little bit about
broader autism phenotype and spend some time trying to understand what might be the earliest signs of autism that are present in babies
way before behavioral symptoms of autism come online. Autism is a complex
neurodevelopmental disorder. It affects in a profound way how children interact with other people. Also produces a series
of very unusual behaviors having to do with
rigidities, repetitive motion and stereotype interest. It’s one of the most prevalent
neurodevelopmental disorders affecting one in 68 children
in general population. Thus far, we really don’t
have biological markers or medical tests that can be performed to determine whether the child does or doesn’t have autism. For that reason, in order
to diagnose the child, we have to perform a close analysis of the child’s symptoms and behaviors. It’s a very complex task,
particularly in young children, because the expression of autism is tremendously heterogeneous. Some children with autism
have tremendous sensitivity around other people
and they are perturbed, sometimes even by proximity of others. Other children are
actually very interested interacting with other kids, but they lack basic knowledge of how to carry these
kinds of interactions in a successful manner. Some children with autism
will never develop speech, others are very talkative or verbose. Some children have tremendous
learning difficulties, whereas others are very
bright and talented in many different ways. So the expression of autism sort of varies along a number of dimensions, and we are still probably to identify other dimensions along
which our children differ. What also complicates diagnostic
differentiations early on is that it’s a developmental disorder. What is means that the symptoms of autism are expressing differently depending on the child’s age and cognitive level. So we can expect quite
different set of symptoms in an 18-monther than in a five-year-old, than in a 16-year-old, and in someone who’s 25. So anytime we see a child
for a differential diagnosis early on, we need to be very mindful. What can we expect from the child, from the developmental standpoint of what kind of symptoms
would be most diagnostic at the given age? What’s important for our discussion today is that autism starts
before the age of three. And this is the period,
developmental period we’ll be focusing on during this talk. So what are the core characteristics of autism in a young baby? Or in a toddler? I think the first thing that we all notice when we talk into the room and the child, they have autism, is that they are not really paying attention to things that otherwise are tremendous important to children without social disability. And these two things are faces and voices. And actually, limited attention to faces and limited orienting
to voice or one’s name is sort of one of the first things that parents begin to notice when they begin to worry about autism. Another thing that’s quite diagnostic and specific to autism is a limited drive to share experiences with others. If you ever interacted with
a 12-monther or 14-monther, you know they cannot do a lot, but they what can do is to pull you into their world of experiences by giving you something,
pointing something, showing you something for no other reasons just for sheer pleasure of joining in the same kind of experience. This kind of set of behaviors
is profoundly lacking in children with autism. Most, or all our children, when we see them for
differential diagnosis have delays in language,
and this is not specific to autism, a lot of children
have delays in language, and they are sometimes
consequential, sometimes not. What’s very specific to autism, that our kids major
difficulties in communicating with other children non-verbally. They are not using their eyes, they are not using their
facial expression and gestures to convey meaning to other
people in the absence of speech. And that affects how they
interact with the world, and in fact, affects quite profoundly what they learn from the world as well. At this age, we are
beginning to see a emergence of speech, but speech
when it comes online, it’s often stereotyped and
not always fully functional. So children not always
use the words they have to communicate in an adaptive way. The speech is often
repetitive, involves naming, repetitive naming of objects around, or repeating verbatim question or phrases by other people. Already in the second year of life, we are beginning to see emergence
of repetitive behaviors, restrictive behaviors
and unusual movements. They’re not horribly,
necessarily, pronounced, and sometimes it’s difficult for someone who’s not well trained in this population to detect them, but we
are beginning to see seeking or avoiding of
certain sensory stimuli, not protecting ears from the sounds or seeking spinning objects. Instead of playing with
toys, with a truck, the child may turn it
around and start moving, flipping the wheels, or begin
to show some unusual movement; it could be just subtle finger
movements or toe walking. Again, not very pronounced but present. So when does autism really emerge? When do we begin to
see behavioral symptoms of autism that we can actually diagnose? This question has been
on our minds for decades, and when I say our minds
I mean our researchers in the field of autism. And we ask this question multiple ways using retrospective parental reports, and more recently, by following up babies who are at risk
for developing autism. And it turns out that in
the first year of life, it’s really hard to see
the kinds of symptoms that we would expect in a toddler. So we cannot sit with the
baby, play with the baby and see, for instance, poor eye contact or some gaze aversion. These behaviors are really
not coming online that early. But by 12 months, we are beginning to see delays in social communication, in use of gestures to communicate. There might be some
withdrawal from enjoyment of social interactions. And parents see it and clinicians see it, so these behaviors are noticeable. The only thing is that they
are not hard with diagnostic. The reason is that some
children who have these difficulties at 12 months
actually grow out of them. And some children who have autism are actually still okay
around 12 months of age. So we are beginning to see
emergence of the symptoms, but they are not yet diagnostic the way we define the symptoms or the way, the things we are looking at. By 18 months, however,
it’s only six months later, we can reliably diagnose about 40 to 50% of children with autism. And that proportion increases
dramatically afterwards. So what does autism might
look like in the first, at 18 months. This is a time when we
say, “Well if the child “seems to have symptoms, the
child probably has autism.” So in order answer this
question, we did something, something very interesting. We pulled data from across several sites in the US and in Canada, and we recruited infant siblings
of children with autism. As you might have heard,
due to genetic liability, these siblings are at enhanced risk of developing autism themselves. So with 700 children,
we had a lot of power to actually ask a question,
what are the unique combination of features
that are predictive of later diagnosis of autism amongst younger siblings of children with autism? And we use a machine learning algorithm, something that we call, it’s called classification
and regression tree approach, which allowed, in which we outputted ratings of the child’s social
and communicative behaviors, repetitive interest on 30 different items. We allowed for all possible
combination of these features to be enter into the model, and then we asked the algorithm
to select those features that are most predictive of later outcome. And we found something really interesting. If you look at this little
tree, this is a little tree, and tells us which
symptoms predict autism. And what we found is
that there’s (murmurs) couple of groups here on the, your right, which sort of were put into this category because they have unusual eye contact, they don’t share or engage others in their world or activities, and they also have very
poor communicative gestures. So those are group of
children with autism, and these are the prototypical symptoms that we looking for at 18 months if we are to look for autism: lack of eye contact, no sharing and poor communicative gestures. So that was great. On the other side, however,
we also have some kids with autism, but they were identified by a combination of different features. These kids had actually
pretty good eye contact, but they were not that bad in
a face-to-face interaction. What they were lacking,
however, is the behaviors that sort of aim to engage other peoples in their activities, and they were showing some emergent repetitive behaviors. So why do we have, why different features predict different subgroups
of kids with autism? Well when we looked
actually at their verbal and nonverbal skills here on the bottom, we found that the children who were showing this atypical
combination of features were actually much higher
functioning than those that were showing a very
prototypical symptoms. What does it mean? This means that our
current diagnostic criteria may be calibrated toward children with greater cognitive
and language delays, and might be missing who
are higher functioning, especially at this very, very young age. Something to keep in mind as clinicians as we are interacting with
children in diagnostic context. The implication of the
study is that right now diagnostic tools and screening tools work on the assumption
that one size fits all. What we are saying is that we
may need to begin to calibrate to increase the precision
of a diagnostic evaluation. We may need to calibrate the screening and diagnostic procedures
to specific subphenotypes within the spectrum. One of the most things
that we were concerned when we started diagnosing
children in such early ages is what is the stability of
the diagnosis at this age. And what we found, and this
is research here at Yale, in England, other colleagues in the US, that follow children
for one year, two years, five years, 15 years, is that the stability of
early diagnosis is quite high. 80 to 90% of children
retained the diagnosis, and those who don’t often have
some residual difficulties or some other related problems. Predictors of later outcome
when we see babies early on are pretty straightforward. Less severe autism symptoms,
higher cognitive skills which enable children to learn faster and greater array of progress over time. What also helps children
in terms of overall outcome is not to have comorbid disorders, and I’m going to talk
about this in one second. So I mentioned at some point a moment ago, I mentioned the special
population of children within general population which are younger siblings of children with autism. Prevalence of autism is
about one in 68 children, in general population. There are certain
subgroups of our population where the prevalence of
autism is much higher. And it’s very important
that, as clinicians, we are aware of that
because that might sort of put a radar on any time we see a child with a specific syndrome or risk factors, that they might be,
that we need to monitor these children also for autism. So one of the groups
that’s at the greatest risk for developing autism are younger siblings of kids with autism. The reason is genetics. And as you might be
aware, genetics factors, are extremely powerful as
etiologic factors in autism with concordance rates for identical twins being around 80% or so. So for siblings who are
not identical genetically, the concordance rates are somewhat lower. And depending on the estimates, they differ between one
in 10 or one in five. So one in 10 or one in
five of younger siblings of children with autism are likely to develop the disorder themselves. Now this number is moderated
by several other factors. One of the factor is how many children with autism there are
already in the family. If there are two children with autism already in the family, the risk doubles. The risk for a younger
child to develop autism goes up to 30, 40%. Another very important factor is the sex or gender of the younger sibling. And you can actually see
it here in this graph, where we have a comparison
of recurrence rates amongst boys and girls as younger sibs. And you can see the boys
are much more likely to develop autism than girls. The risk factor goes up about twice. It’s twice as high in
boys than it is in girl. There’s one more factor
that’s really fascinating, and there’s a lot of
work going on right now trying to explain this, is this gender, sex of the older sibling. So if the older sibling, and
these are the bars right here. If the older sibling
is a girl versus a boy, then the likelihood of the next child having autism is much higher. And it has to do with,
we think it has to do with the genetic load
that girls need to carry in order to reach
affectation status in autism. So this is all new emerging science, and you probably will be
hearing more about that. Important to think about it
when speaking with the parents. Now the siblings, as you often hear, we’re dealing with a
spectrum, autism spectrum, which means at one end of
the spectrum we have autism, and the other end of the
spectrum we have no autism, so to speak, but there’s
everything in between. So when you look at this
figure here on the right, you can see that, sort of metaphorically, that if we think that
there are children here who have autism, younger
siblings who have autism, and then children who
are developing typically, there is a group in the middle, and this is the group we call
broader autism phenotype. These are the children who are at risk for developing some social difficulties, not necessarily meeting
criteria for autism. And they’re also at risk for
developing other difficulties. Already when they are as young
as 18 months or 24 months, we see that there is a
significant proportion of younger siblings of
children with autism who have language delays,
social difficulties, emotional regulation
difficulties, motor difficulties. And what’s very important to keep in mind that when these children grow older, during school age and adolescence, these children are twice
as likely to develop affective disorders, anxiety, ADHD, and a host of other difficulties. So again, the reason
why I’m bringing it up is that we need to be very
sensitive to clinical needs of younger siblings of
children with autism who do not develop autism themselves. And if there are symptoms present, we should be actually acting on them, not worrying is it autism, it’s
not autism, it’s irrelevant. If the child is facing difficulties because of their vulnerabilities,
we need to intervene. It might, in fact,
prevent some of the later mental health disorders that
we see quite so frequently in this population. There are other groups that are also at increased risk for autism,
and one of them are preemies. And as you can see here on the top, the risk for autism increases as the gestational age decreases, which very high risk for autism in babies who are born with gestational
age of 23 or 24 weeks. It’s about 15% or so. Whereas the average for
the entire preemie group is 7% or so. There’s a number of genetic syndromes where we see higher incidence of autism, including Angelman syndrome,
Phelan McDermid syndrome, and more recently,
congenital heart disease. That group is still under investigation. We don’t have a lot of really good studies evaluating prevalence
of autism in this group, but current estimates sort of vary between maybe three, maybe 5% or so. I’m sure we’ll hear more
about it in the near future. Six differences in autism, right? Like many other developmental disorders, boys are more likely to develop autism. If you read DSM-5, the ratio
quoted there is four to one, four boys, one girl. If you look at more recent studies that actually are based on population, they are not as, there may be less bias or more inclusive; the ratio
actually goes a little higher, goes down, and the recent
estimates are just three to one. It’s a very interesting
question, why, right? Why more boys, why fewer girls? When we look at the
severity of autism symptoms between boys and girls, in
terms of social difficulties, the severity’s quite
comparable between the groups. The girls, however, present
themselves somewhat differently because they have typically fewer or not as intense repetitive behaviors, and also they tend to be more introverted or have more anxiety,
have more depression, so they tend to be a little bit quieter. Boys, on the other hand, are very visible, a lot of sir-to-pees, and
they also very rambunctious, very active and prone to
externalizing disorders, including behavioral problems, so they are very, very, very visible. So in our field, again,
this is one (murmurs) areas to watch in the coming years. So we’ve been asking the question, what produces this disparity
between males and females? One set of answers have to do with how we diagnose the kids. Is the current diagnosis sort
of a social construction, or is the disparity due
to biological factors? These two hypotheses, or
these two groups of hypotheses are not mutually exclusive,
they are probably both, most certainly both at play. So let me tell you about the first one. So why do we have fewer girls than boys? Very interesting, several studies evaluated severity of autism symptoms in large population of children,
in England, in the States. And they found that boys and girls may have the same severity
of autism symptoms, but boys are more likely to receive a diagnosis of autism in their community. So there’s something, some
kind of a diagnostic bias that prevents girls who are
clearly socially vulnerable from receiving the diagnosis and being treated appropriately. And the idea is that perhaps
the diagnostic criteria that we are operating within have been created based on
how autism expresses in boys, not how it expresses in girls. So for that reason, it
might not just select those who fit that description. It’s a little circular sort of way of designing the diagnostic
criteria and using them. And this area, again,
is under investigation. We’ll hear more about it
in the very near future. Another biological factors that have been proposed to account for this disparity are heterogeneity, for
instance, in terms of etiology, and that there might be different factors on both genetic and environmental that give rise in males and females. The hypothesis that has
been more strongly supported recently is the hypothesis which suggests that the ratio is due to the fact that the girls may have some
kind of protective factors which are preventing them from expressing symptoms of autism unless they
receive a major genetic hit. So in other words, they need to have more atypicalities in terms of genetics, genetic or epigenetic factors in order to reach affectation status. We don’t know exactly what this
protective factors might be, there are several hypotheses; one suggesting thought there might be some genetic X link factors that are protecting the girls. Or there might be some hormonal factors that also are sort of lessening some of the edges or around
the social difficulties. Speaking of protective factors, we did a study not long ago where actually describably one of the
first, or if not the first, sort of evidence for
potential protective factor already in young babies. This is a study in which we
use eye-tracking technology to evaluate how babies allocate attention in response to a person who
is interacting with them, speaking to them, making eye contact. There’s some toys, you can see the example of a screenshot from that video. So at any given point,
the child needs to make a decision where to move their eyes. And we conducted a study in
six, nine and 12 month-old younger siblings of children with autism. And then we compare how
high risk and low risk, these are the children without
an older sibling with autism, perform on this task. In the red, you see girls; solid line, girls who are
at high risk for autism. In black, you see boys; solid line, boys who
are at risk for autism. And as you can see, something
is really interesting. This red line (murmurs)
is really, really high, both in terms of overall
attention to the task or to the social scene and also in terms of attention to the face of the speaker. It’s a very interesting,
intriguing and puzzling finding. It might suggest that younger sister of children with autism, very early on are employing perhaps a
compensatory strategies to try to get some more
out of social interactions, trying to maybe remedy some of the vulnerabilities they have. And we are following up these studies, and we might have an
answer what it actually is and what it does, but this
is a first illustration of a potential, a protective
factor already in infancy. So just to sum up our
clinical implications of the discussions we just had: early diagnosis is stable, we have tools to diagnose children in a reliable way. We need to be very careful how we do it because we need to
consider multiple factors, not just symptoms of autism,
but also the child’s language, cognition, family
factors, medical factors. But when we take it under consideration and the child’s presentation
is consistent with autism, that diagnosis is relatively stable. There’s no reason to wait for anything. There’s a reason to act and help families move toward implementing
intervention services. Important to keep in mind
that not all toddlers are going to be very prototypical. The girls might look a little different, they might look better,
they might be camouflaging some of their difficulties. Also, kids who are higher functioning might be not quite showing
the prototypical symptoms that we expect at this age. Given the high risk that we know, it’s number of populations,
or subgroups within our population, has
enhanced risk for autism. We have a mandate to actually
monitor these children and screen them to make
sure that we detect social difficulties as soon as possible. And in terms of predictors of outcome, we don’t have tremendously
specific measures, but just on a very general level, having better cognitive
skills, verbal skills and lesser severity,
better progress or response to treatment within the first month typically bode well with later outcomes. All right, so we covered
core symptoms of autism, and I wish this is where we could end. I wish this was all the problems children with autism
and their families had. It’s not true, unfortunately. The problem is that 70%
of children with autism has at least one comorbid condition. And when I say comorbid condition, I mean a diagnosis of
ADHD, affective disorder, anxiety, schizophrenia, many, many, many other disorders. Their prevalence is higher
than in general population. So in some children, or a lot of children actually have more than one
comorbid disorder occurring. What it tells us is that the factors that are driving psychopathology are overlapping to a large extent. But what happens is that
when a child with autism has a comorbid disorder on top of it, it always, it never bodes well
for the long-term outcomes. These children struggle more,
they have more difficulties in adaption, social functioning. They are less likely to get more, they may have trouble getting
an educational placement that’s better matched
with their overall skills. There’s a higher risk of hospitalization, suicidality and many other problems. For many years, these kind of problems are overshadowed by autism. And we typically did not
diagnose comorbid disorders in children with autism, it was like, “Oh, you know, he has
autism so he’s active. “Oh, you know, he’s moody
because he has autism.” DSM-5 changed that, and right now a child with autism can carry
co-occurring conditions as well. What’s important is that,
for us as a clinician, to also keep in mind
that we are not horribly well equipped to differentiate or diagnose these disorders in kids with autism because there’s so much mooshing around of various symptoms and their manifestations. There’s a lot of really, really good work happening right now in our field to create better screening instruments for anxiety, for instance, or
ADHD for children with autism, and also to create treatments
that would be effective in children with autism as much as they’re effective in general population. So again, something to think about, look for, and more information is coming. Why is it important when
we talk about toddlers? Because we are focusing
on kids under three, we typically don’t diagnose anxiety, ADHD, affective disorders in kids under three, so why am I talking about it? Well I am talking about it because we know from extensive research
in general population that there are certain precursors or signs of upcoming difficulties, internalizing, externalizing
problems, already in infancy. And they are conceptualized
under the umbrella of temperament, which is
sort of a biological tendency to respond in certain ways with regard to positive affectivity, joy, or negative affectivity
such as fear and anger, and also ability to regulate emotions that arise in response to
internal or external challenges. What’s interesting that these patterns can be detected as early as six months, and they are relatively stable over time. And there might be shaped
a little bit by experience, but they are relatively stable over time. Extensive research in general population suggests that a typical positive
and negative affectivity are related to later internalizing and externalizing problems. So for us, paying attention
to how our toddlers are doing in terms of their regulation
and emotional expressions in response to real-world challenges is, it makes a lot of sense, and
I’ll tell you why right now. So we evaluated
temperamental characteristics of toddlers with autism. This was probably one
of the largest studies of temperament in young
children with autism. The strength of the study
was not only its size, but we also compared children with autism with children with other
developmental delays. It’s not a big art to compare
autism with typically kids. What we want to see how
children with autism differ from children with
developmental delays. And what we found was
that children with autism seemed to display less
intensely positive emotion, they seem to experience
less positive emotion in response to normative
real-world environment. They also seem to be more
negatively biased early on, and they have difficulties
regulating emotions that arise when they are challenged by new things or things they don’t necessarily like. What was very, very interesting
is that, at this age, severity of these emotional difficulties was not really associated
with severity of autism, which suggests that there may
be totally different domains of functioning that might arise
maybe from different sources and then interact to produce
complex phenotypes later on. We follow this group of
children when they were four, and there was very high stability. You can see these correlations here in the 50s and 60s, very high
stability within our group in terms of what were they temperamental vulnerabilities early on. What was also very important is that this early emotional vulnerabilities contributed later to
severity of autism symptoms as measure one or two years later. So it’s an interesting example,
have two different domains might interact and produce complex, reinforce each other and
sort of produce complex phenotypes in the autism group. So just to sum up, we have, there are very good
reasons why we should be monitoring emotional development
and motion regulation in children with autism. It’s not everything that
you see is part of autism. It might be part of a
different set of risk factors that are likely to
produce symptoms later on that will make their
lives of kids with autism less adaptive, and lives of their families a little bit more complicated. Because of the fact that
the various comorbidities come online at different ages, we might start with some shyness early on, ADHD, anxiety, then we get
into affective disorders, and so forth and so forth. There’s a need of
monitoring kids with autism continuously for these problems. It’s not screening them
at 18 months and say, “Uh, it’s fine.” No, because there are problem
that will arise overtime, so screening kids for these
emotional difficulties throughout their early
childhood and early school age is clinically indicated. And we hope that adequate
or precise detection of these problems and intervention will help improve long-term
outcomes in kids with autism. All right, we’re gonna switch gears now. I’m taking out my clinician hat, and I’m putting on my researcher hat, because we’re going to talk
about the developmental period that sort of spans between
birth and about 12 months or so. And there are very few
symptoms that we can call symptoms of autism during this time. So we are studying this
time in development because we want to understand
what’s really happening. So is it true that autism
really doesn’t start until the second year of life? Or is it true that there
are pathogenic factors which are already unfolding, but they are unfolding on a level that it’s more difficult to
capture with the naked eye? One of the line, a line
of research in my lab is focus in trying to understand why toddlers with autism have
such limited attention, of ability to select for processing, people’s faces, especially
when the people are speaking. It’s a very important, it’s
a core symptom of autism. It’s very important because
we learn by directing our attention toward important
sources of information. In other words, attention gates learning. If we are not paying attention
to people and their faces, we not try to decode what they speak while attending to their facial features, we’re missing a tremendous
amount of information, which is gonna affect not only
how they develop language, but also how they
develop social expertise. And Leo Kanner, this gentleman here, in 1943, he already noted that this kind of limited attention to
people’s faces and voices is present in autism very, very early on. So what we did is we created
an experimental paradigm which allowed us to replicate
what we see in real world, in the lab, and that
replication is important because then we can take it apart. We can manipulate experimentally
what children see, what do they hear, and try to understand what’s generating this profound impairment in selective attention to social targets. And the study involve a video, I mentioned it before, but now I actually will give you some more details. In this video, this very
wonderful young woman is speaking to the camera,
trying to engage the viewer using eye contact, speech. Occasionally, she turns
toward materials on the table, makes a sandwich; occasionally,
the toys start moving. So the child has to decide
millisecond by millisecond where to direct their precious
attentional resources, and where the information
is, what they are going to learn from this scene. When we take group of
typically developing children and show them this video, you can see that they are behaving in
a very orchestrated way. – [Young Woman] Hi, again,
sweetie, how are you doing? Are you having fun? You look so cute today. (rattling and clanking) – So as you can see, all
children kind of converge. When she’s making a sandwich, they observe goal-oriented action. When she speaks, they monitor her face. When something happens
perceptually salient, they look at her face. When we repeated the study
with children who have autism, and you can see the graphs down in red, that this is group of
children with autism; gray, developmental delay; and blue, these are typically developing children. You can see that children with autism are not completely lost. They are behaving quite
well in many respects. However, there’s one situation
in which their behavior diverges from the other groups. This is the situation when
she looks at the camera and speaks, trying to engage
the viewer in the sort of, using prototypical dyadic
bids for engagement. In this situation, they look less at faces than children with
comparable cognitive skills as well as typical controls. Since then, we replicated the study, other people replicated the study. We also extended the
study, trying to understand what are the factors that are separating child with autism from typically or developmentally-late children. And we found that the
factors that are particularly important are the factors that have to do with presence of speech
or combined presence of speech and eye contact. What we also did is we
ask a related question. We know that sensitivity
to faces and speech is present already in newborns. When we show a newborn
a face-like stimulus, or we play human speech, they are capable of orienting preferentially
toward these stimuli. It’s a very elementary
orienting mechanism, subcortically mediated,
something that gives a newborn sort of a leg up to know how to, where to find the most important things that are going to be
relevant to their survival. So what we did is we asked
the following question, we see these deficits
in response to speech and eye contact at 18 months. Can we see them before
behavioral symptoms come online? And we conducted a study
in younger siblings of children with autism,
and as we discussed, about 20% of them are
likely to have autism. Another 30, 40 maybe,
broader autism phenotype. And the remaining children
will be developing typically. And we show them the same kind of task as we did to children
who already have autism. And as you can see, this blue bar is lower than even that of the children
with broader autism phenotype which suggests that,
already at six months, babies who are later become
symptomatic of autism, they already are showing
selective difficulty in attending to faces, especially faces that try to engage them in
a face-to-face interaction. Since then, we replicated
it with a different set of stimuli, and more recently, we actually replicated this phenomenon on a context of a
face-to-face interaction. We took a very, very precise
and very detailed measurement of the child’s attention as
the child was interacting with the person. And what we found was one condition in which children who
later developed autism differed from those who didn’t, and this was the condition when the person sat in front of the child and spoke. Not a condition when she
sang or played peekaboo. It’s only when she spoke, which was quite remarkable
because it really shows the correspondence of
what we see in real world and what we see in the
laboratory conditions. So what we do from here? We are beginning to see
signals as early as six months in babies who later develop autism. These attentional difficulties
are likely to be driven by atypical development of brain networks that are involved in prioritizing
stimuli for processing, selecting certain visual/auditory
inputs over others. These systems develop very early. They develop as early as during pregnancy. So what we are doing now,
we are actually extending downward our studies, and focus on late, prenatal and early postnatal development. What I mean by that is that we are looking for the roots of social and
attentional difficulties that we see at six months already during the last
trimester of pregnancy and shortly after birth. So how do we do that? We do that by conducting
brain imaging studies. And what we are using is the resting state functional connectivity approach. This is a very non-invasive, non-demanding and safe approach to learning about brain
function and brain development. What happens is that in a brain, resting state partners
arise spontaneously, so we can image a child who’s asleep and still draw conclusions
about how the brain is wired to process specific
kinds of information. We can evaluate networks
that are dedicated to processing visual or
auditory information. We can evaluate how the
networks that are involved in selection of stimuli
for processing function very, very early on. The reason why this approach
is likely to be effective, that we know that in older
individuals with autism, some of the networks that I just mentioned are functioning atypically. So what do we know about current, what do we know about current state of network development during the last trimester of pregnancy? It used to be that researchers thought that during the last
three months of pregnancy, what we can see emerging
in the developing brain are largely sensory networks. These are networks that
are involved in motor, in processing visual
stimuli, auditory stimuli, or, for instance, motor function. More recent studies suggest that even before the babies are born,
they are certain foundations that are being laid for
development of brain areas that involved in emotion processing, in processing of speech as well
as social stimuli as faces. After the babies are born
in the neo-natal brains, we are beginning to see the networks that are very important for
us such as saliency network, executive control network,
or default mode networks already form and emerge
and become functional. And we also know that the
properties of the networks are actually linked
with later development. They are linked with language, cognition and emotional development. So what we are targeting right now is something that may actually tell us something about the
mechanisms that underlie early symptoms of autism. And we are tremendously
excited about this work because this is truly
a groundbreaking work, and we are sort of at the forefront of the new research we started. We already started enrolling families. We have over 20 families
already following, which is fantastic. If you hear of any family
interested in participating in our studies, please spread the word. Everything that we do in our work, everything that we do in our clinics, everything that we do in our labs is really designed toward one purpose. And this is sort of my favorite video, and the purpose is to make sure that every child can live
up to their potential at their best, and bring as much joy as we can to their lives. (audience awes) Whoops, so I wanted to thank people, without whom none of the work I talked about would be possible here. We have a clinical team,
we are running a clinic for infants and toddlers who are at risk for developmental disorders, or exhibiting symptoms of
developmental disorders. It’s the Yale Toddler
Developmental Disabilities Clinic. And we also have a research lab called Social and Affective
Neuroscience of Autism Program. Our group is very interdisciplinary, brings experts in multiple areas. And these are people
who are the most amazing people to work with, and I’m truly honored to be able to do that for
so many years here at Yale. So big shoutout to the entire team. (audience applauding) And thanks to the
families who really bring, make everything that we do possible. And our funders who have
to be mentioned as well, they are tremendously important,
influential in our life. So thank you so much for your attention, and I’ll be happy to
take some questions now. (serene twinkling music)

Articles, Blog Tags:

Leave a Reply

Your email address will not be published. Required fields are marked *