Speaking of Psychology - How early detection could change autism diagnosis and intervention, with Geraldine Dawson, PhD

Episode Date: April 5, 2023

About one in 36 children in the U.S. has been diagnosed with autism spectrum disorder. Geraldine Dawson, PhD, of Duke University, discusses why the number of diagnoses has risen so steeply in recent years, why it’s more common in boys than girls, and how research using artificial intelligence and brain biomarkers is making it possible to detect autism risk at younger ages than before – even in infancy. Please help us know more about you and what you would like to hear more of from Speaking of Psychology by filling out our 2023 Audience Survey. For transcripts, links and more information, please visit the Speaking of Psychology Homepage. Learn more about your ad choices. Visit megaphone.fm/adchoices

Before we get started today, I'd like to ask you our listeners to take a moment to fill out our listener's survey. We want to learn more about what you think of this podcast, who you are, and what you'd like to hear from us. So if you could, after you listen to this episode, please go to our website at www. speakingofpsychology.org and look for the link to the listener's survey. We'd really appreciate it if you would take it. Thank you. Now on to the episode. About one in 36 children in the U.S. has been diagnosed with autism spectrum disorder, according to the most recent Centers for Disease Control estimates.

Many of those children are not diagnosed until they're three to five years old, and some even later than that. Scientists used to think that it was not possible to diagnose autism at a much younger age, but recent research using artificial intelligence, brain biomarkers, and other methods has pointed to the possibility of earlier detection and earlier intervention. Today we'll talk about research on autism, autism diagnosis, and interventions. Why has the number of diagnoses risen so steeply in recent years? Why is autism so much more common in boys than in girls?

And how might these new technologies and methods change autism detection and intervention? Welcome to Speaking of Psychology, the flagship podcast of the American Psychological Association, that examines the links between psychological science and everyday life. I'm Kim Mills. My guest today is Dr. Geraldine Dawson, the William Cleland Distinguished Professor of Psychiatry and Behavioral Sciences at Duke University. She is also director of the Duke Center for Autism and Brain Development. Dr. Dawson's work focuses on improving methods for early detection

and intervention for autism and understanding brain function in autism. Earlier in her career, she co-developed the Early Start Denver model, an empirically validated early autism intervention that is used worldwide. She collaborates with colleagues in the departments of computer science and engineering, pediatrics, and biostatistics to develop novel digital health approaches to autism screening and outcome monitoring. Dr. Dawson has won many awards for her work, including a distinguished career award from APA's Society of Clinical, Child and Adolescent Psychop. Dr. Dawson, thank you for joining me today. I'm very happy to be here. Can you start by just giving us some background on what autism spectrum disorder is and encompasses? Because it can look very different in different people. Is that correct?

Yes, you're absolutely right. In fact, there's a saying that says if you've met one person with autism, you've met one person with autism. And the reason for that is that it is a very heterogeneous condition. So, for example, about 30% of people on the autism spectrum never learn to speak. Many may require 24-7 care to help them with their very basic needs. And at the other end of the continuum, there are individuals who are highly intelligent. We have an autistic researcher here at the center who has a PhD in physics. So there's just this huge continuum. So that does beg the question then, you know, what is autism?

What brings all of this together? And in every case, the individual has some challenges in navigating the social world, may have some difficulty in picking up on social cues and interpreting sort of the normal type of social information, such as gestures, facial expressions. That individual also may have trouble forming social relationships. So, for example, developing peer relationships and just understanding sort of the back and forth reciprocal nature of relationships. And then there's also another category of behaviors

which have to do with the tendency to engage in what we call a restricted range of interest, or behaviors. So that individual may be highly interested in one topic and want to talk about that topic and spend most of the time thinking about that topic may also show some repetitive motor behaviors. And then finally, with the DSM-5, an additional trait was added, and that is differences in sensory sensitivity. So it may be sensitivity to sound or to light or to touch. And so all of these things together comprise what we call the autism syndrome. And then, you know, this may be manifest in very different ways depending on each individual. Now, some of our listeners might be familiar with the term Asperger's syndrome, but that term isn't

used now. Can you explain why that's fallen out of favor? Yeah, so Asperger's syndrome was defined as having difficulties in the areas of social interaction and communication, as well as those restricted and repetitive behaviors, but not having intellectual disability or language delay. And so that sort of beg the question of how is that different than just an individual on the autism spectrum, you know, who doesn't have intellectual disability. So, you know, there were many studies that tried to disentangle of those two ways of thinking about people on the highest end of the spectrum. And we really couldn't define any differences.

The other thing is that clinicians were very unreliable in diagnosing, you know, autism without intellectual disability versus Asperger syndrome. And then finally, since many individuals are not diagnosed until later in life, it required the individual to base their symptom or their behavior on history, right? You would have to ask the individual when you were a child, did you have any difficulty developing language? And that was simply unreliable. People couldn't really remember, you know, what their language history was many years later.

So for all those reasons, the DSM Committee, and by the way, that's the American Psychiatric Association Committee, that comes up with the diagnostic criteria, they decided that it was best to go ahead and eliminate that. And think of autism as just a spectrum. And then some individuals also have intellectual disability and language delay. And I think that change has been good.

It did create some, I think, challenges for people who really identified with having the diagnosis Asperger Syndrome, you know, so-called Aspies, or there were a lot of groups that came together to socialize that were diagnosed with Asperger syndrome. And so there was a point made that people could certainly continue to consider that, you know, as a personal way of describing themselves or personal identity. It just wasn't going to be recognized anymore by the, according to the American Psychiatric Association's diagnostic criteria. Well, that makes sense.

Well, let me ask you what we know about the causes of autism. Is it genetic? Is it caused by something that happened to the mother during pregnancy or something else? So the causes of autism are complex. I think it's first, it's important to keep in mind that autism is a condition that is a difference in the way that brain develops. And so just as in neurotypical individuals, there are many factors that contribute to how, you know, each of us has a different brain in terms of its capabilities, its strengths, its weaknesses, and how it develops. And that includes both genetic factors.

So autism is a highly genetic condition. It can run in families. and you do see families with multiple children with a diagnosis of autism. Interestingly, it also runs in families with other conditions such as ADHD. So there seems to be some overlap and genetically between autism and ADHD. So genetic factors are clearly contributing and they are contributing in such a way that the brain is developing differently. we know starting in the prenatal period. And we also know that during the prenatal period,

anything that is going to affect the mother that also affects the brain of the baby that's developing could contribute to differences in brain development and also could, in that sense, contribute to autism. So, for example, one of the theories that is being investigated is what's called the maternal immune activation hypothesis. And what that is is a hypothesis that environmental conditions that activate the mother's immune system will influence the way that the fetal brain is developing. It used to be thought that the brain of the fetus was entirely protected from these environmental conditions that the mother may be experiencing that affected her immune system. But now we know

that actually anything that is activating her immune system during pregnancy could potentially affect, you know, aspects of the brain development, and particularly cells that are called the microglia, which are immune-related cells in the brain that actually participate in the development of synapses and the formation of the brain. So what, what? What can affect the mother's immune system during pregnancy? Really a wide range of conditions. So, for example, stress activates our immune system. Environmental factors such as toxins, so pesticides or air traffic pollution can activate our immune system.

Having a significant infectious disease associated with high fever will obviously affect. the immune system. And in fact, there's been studies now looking at, you know, the impact of having COVID-19 during pregnancy on, you know, the neurodevelopmental outcomes of the baby, and there can be an impact. So we believe that it is this combination of both genetic susceptibilities with environmental factors, particularly those that affect the mother during pregnancies that are shaping all of our brains, right, not just in autism, but also are contributing to differences that we then ultimately would diagnose this as autism. So I mentioned in my introduction, the prevalence of ASD is much higher among boys than it is among girls, four times higher,

in fact. Do researchers know why? Yes, it's been an amazing kind of phenomenon to watch the prevalence of autism increasing over the last several decades. I've been in the field for many decades now. And when I first started as a psychologist, I thought I would be this expert in helping individuals with this very rare condition. And I imagined sort of myself, I was in the Northwest then as being kind of the person that might serve people in the Northwest or the Western part of the United States. And now, of course, you know, autism, spectrum disorder is really relatively common.

And to be truthful, we don't completely understand, you know, even the epidemiologists and experts in the area of, you know, understanding differences in prevalence and so forth, like people at the CDC, they don't fully understand why there's been an increase. But when we do study it, what we find is there really has been changes in both the diagnostic criteria, where they've now broadened somewhat from the criteria that we had early on. And then secondly, we're just so much better at recognizing autism.

And you probably yourself now can recognize autism, whereas 30 years ago you probably couldn't. So I think that we're now able to diagnose autism with individuals who have milder traits and also at a much earlier age. And, you know, keep in mind that these prevalence estimates are really based on just looking at how many times a person is, how many people are diagnosed in a particular geographic area. And so if you see, you know, if we're getting better at early detection, we're going to see these prevalence rates go up. But we don't know whether there might have been, say, environmental factors that also

could have contributed to the increase. And it's also hard to know because the diagnostic criteria have changed over time. So what you're measuring is changing. And that just makes it very hard to know. Well, a lot of your work has focused on early detection and intervention. What does early mean for an autism diagnosis? At what age is it typically diagnosed? Let's start with when it can be reliably diagnosed.

So we know that we can reliably diagnose autism by 18 to 24 months, certainly by 36 months, but 18 to 24 months, that diagnosis is quite stable. The average age of diagnosis, though, in the United States is closer to. five years of age and even higher for families of color or individuals who may not have the resources to, you know, be able to obtain a diagnosis. So it takes many more visits to the doctor, for example, for a black child to receive a diagnosis than it does for a white child. So we know we can diagnose by 18 to 24 months. We're obviously not doing a good enough job. because the average age is much higher.

And by the way, the other thing that contributes to delay in diagnosis is being a girl. Because girls, we are just learning more about autism and girls and finding that the expression of the autism traits is somewhat different. So, as you know, in neurotypical individuals, females tend to be a bit, better at social interaction and picking up on social cues. There's been studies that have shown that, and this is true in autism as well. So you may see better eye contact and therefore many girls are missed. The other thing that contributes to missing children is when autism is combined with a psychiatric condition. So we know that, for example, about 50% of the children,

of individuals with autism also have ADHD. And so what happens is that when a young child presents with both autism and ADHD, the ADHD symptoms tend to get more attention because ADHD is often associated with what we call externalizing behaviors. These are things like, you know, you can't sit in your seat. You might be interrupting. You can't finish things. You might be, you know, blurting out.

And those behaviors are very disruptive and they tend to catch our attention more. And we call this diagnostic overshadowing when a person has two conditions and one is kind of more obvious. That diagnosis tends to get the label for the child and then the other one is missed. So when children have both autism and ADHD, they tend to be diagnosed at a much higher age. So, but I think the second question is, you know, when do the early behaviors or early signs of autism actually begin to emerge and how early might we be able to detect autism? So there's a lot of work going on and we've been doing some of this work in our lab as well, trying to identify autism in babies because there's been many studies now that, and some of this work was done by our team where we looked at home videos of infants who then were later diagnosed of autism. But there's been many studies that

have shown that between about six and 12 months of age, that the signs of autism begin to emerge in the majority of infants. And those signs could be the lack of communicative babbling. So we know, you know, during that period, babies tend to use consonant vowel sounds, you know, mama, blah, and they engage in this sort of back and forth with their parents. That is often absent or much delayed. The use of gestures, so we know that during that period, babies begin to start to use simple gestures such as bye-bye. They also develop what is called joint attention, which is a set of gestures where the baby is drawing the parents' attention to something, and they could do that, for example, by pointing at something to show them something,

or they might be interested. If someone else points, they would follow that point and notice what others are doing. That's why we call it shared or joint attention. And those behaviors are quite diagnostic. And so we see this lack of pointing, particularly to show. So another sign could be some lack of interest in engaging with people as compared to objects. So this baby may be very interested in objects and features of the non-social environment, but maybe making less eye contact and not playing those baby games that we often see like pat a cake.

And so those and even the sensory sensitivities have been. noticed in these young babies. So we should be able to pick up on autism during the infant period, and we're working on a lot of different ways that we may be able to do that. Some of those are ways of measuring behaviors. Other ways are trying to see whether we could detect early changes in the way the brain is working and functioning, and I'm happy to talk about those. but there's a lot of work on looking at these brain-based, but are called biomarkers, to see whether we can detect these even before we might be able to diagnose the full syndrome, which would be, you know, not until, say, 16 to 18 or 24 months of age.

When you're looking for brain-based biomarkers in very young children, how do you do that in a lab? I mean, how do you take a baby that's six months old and do, you know, an fMRI, for example? You can hardly do that with adults sometimes. Actually, babies are easier than adults. Yes, for us that do brain research, it's interesting that babies, they take naps and they, you know, they are actually much easier than the hardest period in terms of some of these measures is the toddler period where they want to, you know, get up and wiggle around.

But there are lots of different ways. of studying the brain in infants. And most of the studies to date have followed infants who have an older sibling with autism that's already been diagnosed because we know that when there's one child in the family that's been diagnosed that the probability that the second child will have autism is about one in five. So that is a much higher rate. And so by following these what are called high likelihood infants prospectively and then taking brain measures and behavior measures during infants, infancy, we can then follow them longitudinally.

And then when that one in five is diagnosed with autism, we can look backwards and say, oh, well, what was different about their brain when we made those early measures? So that's the ways that the studies are mostly done. And in terms of the measures, there's been quite a bit of research looking at the development of the structure of the brain using MRI. So these are large collaborative studies occurring at multiple universities. In fact, one of the biggest studies is led by my colleague Joseph Piven, who's a psychiatrist down the road. at UNC, Chapel Hill, because I'm at Duke, not very close by. And so he, for the last several years, has been falling very large samples of high-risk infants or high- likelihood infants and then doing repeat MRIs to watch how the brain

develops. And what he's found, he and his team, is that between six and 12 months of age that the white matter of the brain. So this is the part of the brain that helps different parts of the brain communicate with each other. So this is the myelin that goes around the connections of the brain that makes it more efficient in terms of the ability to communicate between two brain regions. And what they found is that there's differences in the development of this white matter that can be seen by six to 12 months of age. And essentially, what people are finding is that autism is a condition that affects the

neuroconnectivity across different regions of the brain and these white matter fiber tracks. And if you think about the behaviors that are affected by autism, which mostly is a around social interaction and communication. Those behaviors require the precise coordination across multiple brain regions. So think about, for example, if you're involved in, say, a conversation with another person, you're watching their facial gestures,

you're listening to the sound of their voice, you're anticipating the timing of the back and forth of the conversation, you're using your motor cortex, to say words, for example. So even something like a conversation requires, you know, the frontal lobe, the visual cortex, the speech areas in auditory cortex, the motor cortex. And these all have to be functioning in precise coordination.

And so we do think that autism affects the ability of the different brain regions to be coordinated in a way to allow us to. engage in these very complex behaviors. Now, I do want to take the opportunity here, though, to stress that autism also is associated with many positive traits, skills, and talents, and that it's, you know, very likely that the sort of sacrifice of the ability to do some of these, you know, behaviors that are important, obviously social interaction and communication. but it also offers the opportunity to be able to really view the world in a very different and unique way

and to have very important and special talents. And it probably has to do with the fact that there's probably large areas of the brain that are now able to be used for other things, you know, whether it's mathematics or music or art or physics. You know, so, you know, it's important to. to really think about autism is a brain difference that then comes with challenges, to be sure, but also with many very important and special strengths and talents as well. Now, you published a study in February that found that an artificial intelligence algorithm could detect autism risk by screening infant's medical records right from the first month of life.

that's pretty incredible. How does that work? So we have been, we were awarded a grant from the National Institutes of Health that challenged us to come up with ways of screening for autism in babies. Because as I said before, it's clear that autism starts in infancy. And we want to begin, you know, intervention, services and support is as early as possible. possible. And so we actually have been attacking this two ways. So one is to gather information about the baby's health care record to see if there are clues in terms of how the infant is utilizing the health care system that might indicate that that baby would be more likely

to have a later diagnosis of autism. And the reason why we thought this might work, is because autism is associated with quite a few medical conditions. So people think about autism as a behavioral health condition, and it is, but it's also a physical health condition. So, for example, many people diagnosed with autism also have difficulties with the GI system. So in babies, this could be reflux, it could be diarrhea. constipation, other GI-related conditions. And actually in our studies, we are finding that these start very early within the first three months of life. So this is happening before we even start to see changes in behavior.

A second one is differences in the way the motor system is developed. So we find that even during the first year of life, that babies who are going to go on to have a diagnosis of autism are visiting the physical therapist for difficulties in their motor development that are affected in the first year. We also see higher rates of sleep disruptions, trouble establishing a circadian rhythm. And in fact, there's been research that shows that some of the genetics, that genetic genes that are responsible for regulating our circadian rhythm. may contribute to autism. So there are other conditions as well,

but what we found was that if we can study how the parent is bringing their baby to receive these medical services in the first year of life, that we can do a pretty good job of predicting, even in the first month, that this baby will have a higher likelihood of a diagnosis of autism. And the idea here is that if, you know, there's no way that a pediatrician could be, well, it would be very difficult for the pediatrician to have to look through and try to understand all of these patterns of health care utilization when the pediatricians say seeing the baby for a well-child visit. So, but the computer can do this, right? So we can have the computer gathering all of this health-related data.

It could be all the diagnostic codes, all the types of visits, all the medications, all the procedures, and then use artificial intelligence or machine learning to combine all of these data into an algorithm that then is predictive of a particular diagnosis, in this case, autism. And we've been looking at the same thing for ADHD, by the way, and we find that these can be distinguished. So, for example, a baby who is going to have a later diagnosis of autism is going to be much more likely to have visited the emergency room by age one. And this may have to do, again, with that impulsive behavior, risk-taking, you know, other things that might lead to accidents, right, that would end up in an emergency room. So, you know, there's just a lot of information there that that then can. you know, help us to let the pediatrician know that they need to be paying special attention to this baby and monitoring the baby's development very closely.

And this information can be combined with what is called clinical decision support. So this is a very active field now in healthcare where pediatricians and other providers are provided with information about whether or you. this patient, you know, is a higher likelihood of developing, you know, a condition. It could be a heart-related condition, in this case, autism. And then also providing the physician with, well, what should they do about it? So giving the physician through the computer advice about, okay, this means that, you know, you may want to make a referral for such and such, or depending on the age, you may actually want to refer for a diagnostic evaluation. But, you know,

you know, here's ways that you can follow this baby and screen, you know, to see if the behaviors are merging. So that was one approach. And by the way, that work was led by my colleague Matthew Englehard, who is both a physician and a biostatistician. So he combines both medicine and data science to do this really phenomenal work. And we're now, we've received more funding to continue to expand upon. this work. And then the second approach that we've been using is to use the computer to measure

those early behavioral signs. So this work began to try to improve our ways of screening for autism. So currently, the American Academy of Pediatrics recommends that all toddlers be screened universally for autism at their 18 and 24 month visit. And remember, that's the age where we can make a diagnosis. So, you know, it's good in terms of you can have, you know, a diagnostic evaluation at that age. But it's not so good in the sense that what happens is that they often get screened. And then it takes time to make the referral. Then they get on a waiting list for a diagnosis.

and there's often a long waiting list, and then after they get the diagnosis, they're on another waiting list to get the intervention. So, you know, there can be a year or before intervention begins. So now we're looking at 36, you know, or even, you know, 48 months of age. So we really want to shift all of that earlier. But there was problems with this screening approach.

So typically how this is done at 18. in 24 months is with a parent questionnaire. And it's a simple questionnaire given universally, for example, here at Duke in primary care, where you ask the parents about 20 questions that have to do with those behaviors that I mentioned earlier. Is your child pointing? Are they using words? Are they interested in games?

Do they imitate you? So questions like that. And then depending on, you know, how many yeses, or knows in this. case that the parent says, then they would be considered to have a higher likelihood of autism, and hopefully the provider would make a referral. Now, one of the interesting findings is that most pediatricians still don't make a referral. About 60% of the time, they, even with a positive screen on a parent questionnaire, they don't make a referral. They take a wait and see approach. And part of it has to do with just their lack of confidence and interpreting the information. And part of it

may have to do with the difficulty in finding services and some reluctance to, you know, send the parent on a journey that that can be very frustrating and costly. Or telling the parent that they think that the child is on the spectrum and then maybe being wrong, right? I mean, they could be. Yes, absolutely. Yeah. And I think that's an important point too, because so the most common questionnaire is called the modified checklist for autism and toddlers or the mchat. And by the way, we are so happy we have the M-chat and we are so happy that pediatricians use it. So I'm not in any way wanting to say we should discourage use of the M-chat. But if the child is referred, they will have about a 50-50 chance of having a diagnosis of autism.

So that does mean that there will be some, you know, false positives, many people waiting to get a diagnosis that don't need to and adding to those long wait lists. So all of that, I think, is contributing to this reluctance to make a referral. But there were other problems, or there are other problems as well. So the M-Chat does not work as well with families of color. So we don't completely understand it, but there are. reasons why from different cultures that have different attitudes around providing information or how they interpret information. The other thing is that it does require literacy and understanding of child development. And, you know, many parents may not have those skills. And so for those reasons,

when it's used in real world settings like primary care, it's really not performing as well as it should. It also is not as good at picking up autism and girls. So what we wanted to do was to develop a methodology where we could directly observe the child's behavior. Because we know, I mean, autism is a behavioral diagnosis. So we should be able to directly observe what the child is doing and try to pick up on the behavioral signs. And some people have tried to have the pediatrician do this, but honestly, they don't have the time. And also, they have so many things that they're screening for and concerned about. They don't have a time to do a diagnostic assessment as part of their work. while child 15-minute off and visit.

So we decided to try to use a technique called computer vision analysis, which is a way of the computer with a computer, using a computer to measure behavior. And the way this works is we designed an app, and this can be downloaded on a smartphone or a tablet. And the app displays a set of short. movies. And these movies were strategically designed by our team to elicit autism-related behaviors. So to elicit facial expressions and vocalizations, to assess, to be able to see whether

the baby is more interested in looking at objects versus people. And so these, again, were very carefully designed. So the baby watches these brief movies. It's all done within less than 10 minutes. And the camera that's in the device, whether it's a smartphone or a tablet, is recording the baby's responses. Then the engineering team uses computer vision to automatically code a wide range of behaviors.

So a computer can code, facial expressions, vocalizations, head turns, body movements. You can do all of this. And what's really amazing is it can do it better than a human. And I truly believe this. And I, of course, have been observing behavior as a psychologist my entire career and coding a lot of behavior. But I'll give you an example. So one of the early signs is not orienting when your name is called.

So while the baby is watching the movies, someone behind them actually calls their name. And, you know, the baby will typically turn their head. And a computer can precisely measure that. And what we found was that, indeed, that babies who then later had a diagnosis of autism did not turn their head as frequently. But some of them did turn their head. And what we found was that when they turned their head, they did a second delayed compared to the neurotypical babies. And there's no way a clinician could have seen that.

So the clinician would have said the baby doesn't have that sign, but the computer was able to do this at a much higher resolution. So we have now validated this in the 18 to 24 month age. And we also have just completed a study where we've shown that this can be done at home. So parents can download the app at home and those data can be sent to the pediatrician ahead of time right before the visit so that they have. have assessments of the child behaviors that could also be done over time to track behavioral development. And the study we're doing now is starting at six months of age and then following babies longitudinally and they were repeating the app over time and they will see whether this can be used to eventually pick up on signs in, you know, in babies who then will have a diagnosis

of autism. So I think it's, you know, I started. my career by trying to understand autism and infants by looking at home videotapes that parents had gathered. And now I'm using artificial intelligence and computer science to measure these on an app that's on a smartphone. So I would have never imagined that. Technology. You just can't keep up with it. It's amazing. So let me ask you one wrap-up question, which has to do with why the early diagnosis is important. And obviously, we've been talking about the concept of interventions of that I actually describing them. But if a parent gets the diagnosis early enough, what is it that the parent should then do to help the child? Yeah, thank you for that question, because that is,

you know, really the crux of the matter. The reason why we want to detect autism early is to provide services and intervention. So I'll describe a type of intervention that is very commonly used now. These are a class of interventions called naturalistic developmental behavioral interventions. They evolved out of traditional applied behavior analysis, which is a technique that has been used for many years as an intervention model. This is sort of a newer version that is much more play-based, much more naturalistic. We follow the infant or the toddler's lead and capitalize on their interests and preferences, which makes the therapy much more engaging, both for the child and the parent. But essentially, these interventions use play and naturalistic interventions

to facilitate the babies, first of all, interest in other people. So to bring the baby's attention back to people. Because remember, this infant or toddler is now much more interested in toys and the world of objects than people. But there are ways of using play and inserting yourself into the interests of the baby or the child where you become part of their world. and then they will naturally start to be interested in you and looking at you. Once that happens, then we again use these techniques to teach the infant or the toddler how to use gestures, how to communicate through a set of strategies that can be taught

to parents to use at home or, you know, delivered by a therapist. And so these parent, what are called parent mediated therapies are ones where we work with the parent to learn these strategies that they can use during their everyday activity. So if you think about a neurotypical infant, we don't sit them down and say, I'm going to teach you now how to speak or I'm going to teach you how to recognize that this is a smile. it happens during every interaction during the day is an opportunity for learning during meal time and bath time at the park. And so similarly, these interventions that are taught to parents are ones where the parents

can use these strategies just throughout their everyday activities. So we have a book, for example, it's called an early start for your child with autism, and it is written for parents and it teaches them a wide range of strategies they can use. And it has to do with things like how do you position yourself. For example, when you sit down to read a book with your baby or your child, you're not going to sit them on your lap, you know, where they're faced away from you. You're going to sit them where they're faced forward to you. So you're going to have yourself in their zone of proximity.

and also in their spotlight, so to speak. Then you're going to learn how to kind of capitalize on the things that the babies is interested in to start to facilitate things like turn-taking. You're going to do a lot of imitation of the babies or the toddler's vocalizations and actions. So we find that, let's say that this two-year-old is banging on something and not looking, at you. If you have a drum two and you bang at the second same time, that two-year-old with autism is going to be start looking at you, playing with it. And pretty soon, you're in a fun back-and-forth game where you're both, you know, playing the jumps together. And so these are

relatively easy to learn. Parents are very good at learning them. And they've been shown to really facilitate early social and communication development. So we're very excited about, you know, parents using these. And again, they can be delivered by therapists as well. And they can make a huge difference in children's outcomes. Well, Dr. Dawson, I want to thank you for joining me. This has been really interesting. And I'm sure that our listeners are going to learn a lot from hearing from you. So thank you. Oh, it's been my pleasure. You can find previous episodes of Speaking of Psychology on our website at www.w.combeingof Psychology.org or on Apple, Stitcher, YouTube, or wherever you get your podcasts.

And again, we'd like to hear from you about what you think about this podcast and what you'd like to hear from us. So please go to our website, speakingof psychology.org, and look for a link to the listener's survey. If you have comments or ideas for future podcasts, you can email us at speakingof psychology at APA. Speaking of psychology is produced by Lee Winerman. Our sound editor is Chris Kondyen. Thank you for listening. For the American Psychological Association, I'm Kim Mills.