How to Talk to Children with Autism and Developmental Disabilities

(whooshing tones) – Okay well, it'swonderful to be back here.

It's exciting to have anew venue and to know that we don't have to come to a screeching halt when buildings changeand I'm really excited that the audience stayed intact even though we're in thispublic health emergency.

So thanks everybody for being here.

No financial disclosures.

So, I have been working in thearea of language research now for several decades, I've really seen lots and lots of circumstances wherechildren have challenges in learning language.

I've worked with kids who have deafness and children who have hearing loss.

And I'm getting echoes is that a problem.

Yeah, all right we'll go back.

Okay, we'll go back here, is that still okay in the back? Yeah.

But there have been some unique situations that I've had workingwith children with autism and I wanted to tell you one story to set the stage for this talk.

So, I was in an inclusion classroom, there were childrenabout maybe six children sitting around a table, andthe teacher was going to do a game with colored ballsof clay and sharing.

So she reached into a bucketthat had these colored balls of clay and she pulled out two, and said to child number one, “Which one do you want, theblue one or the pink one?” And the child one went, pink, and she said, “Pink it is.

“, And she handed the child the ball.

She reached into the bucket and pulled out two for the second child.

She said, child number two, “Which one do you want? “The green one or the or the black one?” And the child said, “Green.

” She said, “Green it is.

” And she gave the child the green ball.

She reached into the bucket, she pulled out two more, and she said, child number three, “Which one do you want, thered one or the orange one?” The child went red, and sheis about to give the child the red one, when a hand from an adult who I hadn't even been paying attention to came across the child likethis, blocking his mouth.

And this adult said, “I”.

And the child's whole affect change, and the child goes, “I”.

And the adult says, “Want”, and the child repeats want, and the adults says, “Whatcolor did I say, red?? And the child said, “Red.

” And I was stunned that wasthe way that this adult, well meaning, well educated, prepared, thought that you shouldteach children language, if those children had autism.

And this bothered me forseveral years and recently I just felt like I needed to move forward and understand whether my instincts, that wasn't such a goodtechnique, were right.

I had inspirations in clinic, because I would see thatsame kind of prompting and excessive repetitionof these stock phrases and short sentences, andI would try interacting with the children and I could see them begin to learn sometimes inour very assessment session.

So I decided I needed to do something and that was reallylooking to the literature and maybe delve into doing some change in how we teach childrenwith autism, language.

So with that in mind hereare the learning objectives for this talk.

By the conclusion of the talk, I hope that you're allgoing to be able to define language nutrition anddiscuss the components of it.

Discuss methods for assessing the language environment of children.

Contrast language nutritionfor children who are typically developingand children with autism or other developmental disabilities.

Evaluate the impact oflanguage nutritional and language development in autism, and then have a start atleast at advising parents of children with autismabout how they can talk to their children.

Here is the take home message.

Nutrition is very importantfor physical development and health, languagenutrition is very important for language developmentand language health and languages nutrition is as important for language health andlanguage development with children with autism, as it is with healthytypically developing children.

Okay, so now let's seeif I can convince you of those take home messages.

So first of all, letme tell you what I mean by language health.

So the evolutionary psychologists tell us that people been talkingfor someplace between 70 and 200, 000 years, and it is pretty clear that language has been very important in how the human society has taken shape.

We cannot only think about hypothetical, or think about the future, we can talk to others abouthypothetical and the future.

And in so doing we can beginplanning and organizing.

In the old days, that mightbe to kill the woolly mammoth, and these days it mightbe to put men on the moon.

So, this has always been really important.

At the same time, there'salways been options for people if language was not astrong suit of theirs.

You could work as afarmer, or as a sailor, or as a carpenter, and youcould be very, very successful.

But as we enter the 21st century that's really beginning to change.

So sailors need to learn how to use GPS, and farmers too.

And they learn how they can differentially water their crops for a better yield.

And carpenters may havemore sophisticated tools with computer generation, and so they're going to have to read manuals andmaybe even change programs.

So language now in the 21st century has taken on a much more important role.

It's becoming much moreimportant in education, literacy, which is anoutgrowth of oral language, is so critical to success in school, and it's also really importantto success in an occupation.

And that is true forindividuals with disabilities as it is with individualswith no apparent disabilities.

So the job options for an individual with autism are far greater, in fact, if that individual is a competent communicatorand knows language, then for a person who remains non-verbal.

So we know that thepath to language health begins very early in life, and there is a sequence.

Here it is kind ofschematically organized.

There are certain things that come early like recognizing names.

Then there's gestures, children begin to communicate with their hands and their body.

Sometimes usually aroundthe first year there's words and then the vocabulary grows, then words come togetherand then pretty soon children can stringtogether many sentences and carry on conversations and then use this language for reading.

Now what's important tonotice on this slide, is not just the sequence, but each of the bars is quite stretched, there's a range, there's variation in how quickly, children learn these skillsand when the skill comes under their control in their lifespan.

And you know it's a littlebit like height curves, you can look at curvesat how quickly children do some of this and you'll find that in cross sectional data, there are quick learners, and less quick learners.

So, this is a slide aboutthe size of vocabulary is measured by apparent report measure, called the MacArthur-Bates, and this had about almost 5, 000 children, feeding this data set.

And you can see that the children in the kind of pinkish colorare the ones contributed to the fast path up there.

And the ones with purple were contributing to the lower 10th percentile.

Wide variation in the rate at which children learned language.

And like the heightcurve this was generated on cross sectional data butit can also be pretty useful for longitudinal data.

So these are longitudinal data, slightly different method, but child age on the x axis and number of differentwords that were recorded during a laboratory session on the y axis, and you can see that thereare children like this who are quick off the blockand stay ahead of the pack.

And then there are childrenwho are slow developers, slow to get going, and then take off.

So there's a wide variation in how quickly children learn this important skill, that's taking them toward language health.

Now, what contributes to the variation in the rate of development of language? Well, there's lots of things.

The child, of course, bring something to the language learning situation.

Certain children havemore challenges based on, we're not exactly sure what.

Maybe structures orconnectivity in the brain.

But there are a lot of factors that also come from the environment thatwe know have been associated with rates of development, things like caregiver education, family income, and alsomaternal depression.

And we know that these thingsare all interconnected.

So, caregivers who have a good education are often able to get better jobs, they can bring in more income and live in communitieswith better schools and create a healthier maybeliteracy rich home environment.

Now I want to now turnfrom language health to language nutrition.

So, what all of thosedifferent factors boil down to, is what is the languageenvironment for the child who is learning language? And the term languagenutrition brings together, other terms that have beenused in the literature, one has been calledbaby talk and that term usually relates to the very early stages of talking to children, where adults and otherchildren for the record, use exaggerated intonation, and slower pace, and a slightly modified vocabulary in order to gain the child's attention and also to promote their learning.

Infant-directed speech isa little bit more general, and includes other featuresthat go into the child, into this speech thatparents use with infants, of course, the customizing to the learner doesn't stop with infancy.

So, another term in the literatureis child-directed speech and sometimes it's just calledinput or environmental input.

I want you to think of all of those as the language nutrition for a child.

This idea that the language environment or language nutrition isimportant for language health and language developmentwas really promoted by a very influential book called, “Meaningful Differences”, published by Hart and Risley in 1995.

And what these two researchers did, was they took a hugesample of 42 children, they were carefully selected so that they were a certain age, they were between seven andnine months in enrollment, and they were selectedso they would probably be available for the fullthree years of this study.

And they were also carefullyselected to represent different socio-economic strata.

So there were upper-middle, upper-class, middle-class, lower-class, and families on welfare.

And these researchersput a recording system on the dining room table and collected about an hour a day of input, and then transcribed all that, and began trying to understandvariation in the environment, and then also tested thechildren at age three, to look at variation in the child outcome.

And so the findings, the first thing is almosteverything a kid says, was available on the tapesthat the researchers collected.

So that is the kind of clear evidence that children are learning, primarily from the people around them.

There were dramatic differences in the size of thevocabulary at age three, not unlike, even in this small sample, not unlike the slides I had shown you.

There were dramatic differencesin the parent input.

There was a very strongassociation between how many words parents presented to the children around the dining room table, and how much the child hadlearned by three years of age.

And the sad fact was that the middle-class and upper-middle class families presented many morewords to their children and the children were much more successful in learning language.

And by extrapolating fromone hour a day recording to a 14 hour a day being awake, to seven days a week, to three years, the researchers established that there was a 30 million word gap between the highest SES familiesand the welfare families, and that was a reallyshocking demonstration of how vastly differentthe language environment, or the language nutritionfor children can be.

Now if you go to literaturenow and you Google, Hart and Risley and 30 million word gap, you'll see that there's a big debate.

It's not 30 words, it's six million words.

The basic fact that certain families present a very rich language nutrition and other families a lessrich language nutrition has not really been challenged.

So let's think, what composes languagenutrition environment? What comprises, there you go.

Okay, so the three elementsto language nutrition that I want to emphasize.

The first is the quantity of the language.

That's what we've just been talking about with Hart and Risley.

How many words are therein the environment? So it's total number of words and their strong associationof language outcome like we just talked about.

There's also gestures, which can be extremely communicative.

So that goes under the quantity.

The second element is quality, and this is quality ofthe linguistic input, or quality of the language nutrition.

And this could be thenumber of different words that families use.

The sentence complexity, the number of questions, like who, what, where questions.

And talking not only about the present, but the past, and the future.

So those get lumped underthe quality of the language.

And the last componentof language nutrition is the quality of the verbal interaction.

So these words andsentences can be delivered in ways that are warm andresponsive to the child that are expansions fromwhat the child already said, or they can be offered in waysthat are more adult driven, and maybe come out ofthe blue for the child, so they can reflect a jointengagement about a single topic, or they can be dispersedand all over the place.

Sometimes parents are quite intrusive, They're the ones whogenerate what the next topic of conversation will be or very directive, telling the child what todo, pick this up, do this.

So all of these representcomponents of language nutrition.

So now I might need some helpgetting this video to run.

So let's take a look at this.

Some of you may havealready seen this video, let's take a look at this video, as an example of language nutrition.

Let it rip.

Needs sound.

– Did you understand it though? – No.

– No.

Okay, all right.

(baby babbles) Huh? Oh, nah, not this one, this the grand finale of this one.

(baby babbles) Yeah that's the next one.

(baby babbles) That's what I was wondering.

I didn't know what they'regonna do next season 'cause they did some stuff.

(baby babbles) Exactly what I was thinking.

(baby babbles) Oh yeah, yeah.

(baby babbles) Right, don't bring that again.

You know what I'm sayin', didn't do the same stuff, you know what I'm sayin'? Yeah, it's like that, yeah.

(baby babbles) Yeah, like go somewhere else with that but don't right here, you know what I'm sayin'? Yeah.

(laughing) (baby babbles) That's what I'm saying'.

It was like (exclaiming) you know what I'm sayin'? I was like what in the world? Don't do that here, youknow what I'm sayin'? Yeah.

(baby babbles)(audience laughs) Really? I thought the same thing.

(laughing) We think a lot alike huh? – Okay.

(baby babbles) – That's crazy.

(baby babbles) Right.

– Okay.

(audience claps and laughs) Okay one or two comments.

You're all smiling and warm, feeling cozy.

Tell me one or two thingsthat you really loved about that video, just scream 'em out.

– [Woman] Warm, warm.

– Warm, and responsive.

Yeah, what else? – [Woman] Fun.

– Fun, really great.

– Interactive.

– Interactive, right.

Lots of gestures, verbal andgestures putting together.

Kid driven right, the kid said, “Yeah, I think the same, we think alike don't we?” So that is I think it's likea quintessential example of language nutrition.

Now, just want to be clear that language learningopportunities vary across the day.

You have opportunities in the morning when kids are gettingup or getting dressed, and they go all throughthe day even until bedtime.

And the way we used to assess this was some kind of old school methodology like a tape recorder, or handwritten notes when children came into the laboratory.

This is my old research.

I actually used video, but I just want you to know that there's some newschool ways of doing this and one is called the LENA device and that's the picture here on the right.

And this is a specialrecorder with an algorithm that gives you approximate word counts for language it's nearand close to the child, and the child wears itin specialized clothing you can get about 16hours of home recording.

They really get a feelfor what's happening in the child's natural environment, and there's lots that we're learning about language nutrition whenwe see the outputs of this.

So here's hours of theday across the bottom.

And this is the adult word count in 10 minute little dollops, and you can see that in themorning there around breakfast and probably maybe around school, there's lots and lotsof input to the child, and then you see that it drops down between 12 and two as probably nap right, and then it goes up, andthen there's a big old peak close to 8:00 P.

M.

Now if you look at anotheroutput from the same LENA device.

This one is called conversational turns, and this is when the child and the parent go back and forth, and back and forth.

And you can see in that morning segment, there's lots of back andforth, and back and forth, with the nap there'snothing back and forth.

And then at 8:00 P.

M.

notice, there's not a lot of backand forth there either.

That's probably because that's the reading and I'd say the adulttaking a little bit more responsibility for the language input.

So there's lots of interesting things we can do with a LENA, and I should tell you that some speech and language pathologistsare using this now as a clinical tool andthere's ways that families can also use it to seeabout language environment, let's say for example if they'renot home with their child.

So what I want you to knowis that language nutrition has been found to be a keypredictor of language outcomes in a wide variety of studies and a wide group of children, different language groups, children who were born preterm, children who have a hearingloss, Down syndrome, and specific language impairment.

So now let's start thinkingabout how this might work in the circumstance ofchildren with autism.

We know that this has really captured our national attention.

But I think this aspect oflanguage learning has not.

So the first thing isI want to convince you that there is variation inthe rates of language learning in individuals with autism.

Now I know this groupisn't so surprised by that.

But I just want to make the point that there are different curves, as you go across time andthis is a very wide age span.

From infancy all the way to adulthood.

The red is the normativeor typical development, and you can see some individuals have very close to typicaldevelopmental status and some have a more marked delay.

I think the other thingthat you can see is that the more of the variation is taking place in those early years, let'ssay, around two to three and by the time kids aregetting to be about six their rate of developmentdoesn't change very much.

Now that is not to say thattheir language doesn't change, you'll notice that they still keep getting a higher age equivalentwhen using the Vineland, which is this study, but theirrate of change gets stable.

So if we're gonna do anythingwith language nutrition best hunches we want to do it when there's a lot ofvariation and movement, and that would be in the early years.

So I worked last year on a scoping review.

I learned about a scoping review from my wonderful colleagues.

And this is an opportunity to look broadly across the literature, not necessarily to reduceeverything to a systematic review, but to look for trends and truths in the literature across a wide range, and we did a scoping reviewbecause we wanted to see how language nutritionvaried in different groups and we chose three groups, children born preterm, children with intellectual disability, and children with autism.

So I want to just giveyou one other kind of set of basic concepts beforeI tell you what we learned in the literature.

So you can think about languagenutrition as prevention.

Prevention of the moresevere language disorder.

But I wanted to remind youthat there's three levels of prevention in a public health approach.

The first is primary prevention where you give the treatment to an asymptomatic individualand they never get the disease and of course for thoseof us in pediatrics who know the quintessentialexample of that is vaccination.

Ideally you give measles vaccine and the child never gets measles.

But there's also secondaryand tertiary prevention and you may rememberthat secondary prevention is that you give thetreatment to an individual at an early stage, and thenthat individual develops a very mild form of the condition, and that's an example of like mammography.

You know we wait and we get mammography, but we're not anticipatinglike the vaccine.

When we see a mass we have to act on it, and that leads to better outcomes.

And tertiary prevention isthat we know the individual has a condition, thiscase it could be autism, but they get to treatment, and they have better functional outcome.

And I think thequintessential example of this is early intervention.

So if we have an individualwho has Down syndrome, or other intellectual disability, and we give them early intervention, we know that they may stillhave cognitive impairment, but we really think thatthey'll be more successful functionally than if theyhadn't had early intervention.

So when we thinkingabout language nutrition, I think we're sort of focusing, especially for autism, we're focusing on secondary and tertiary preventionand not primary prevention.

So now I want to show you a little snippet of a father and son.

The boy is about five and has autism.

So let's see if this works.

– What else is in this box Danny? Take another toy out.

– Duck.

– Quack with a duck.

Big bird, does he look like a duck to you? – Bird.

Can you make the bird standup? – Yeah.

– Atta boy.

Okay, what else is in that box? Show me something else, you have another toy in there you want to play with? What's that? What's that toy? What's that called? – A truck.

– That's a truck.

You like that truck?- Yup.

– What color is that truck?- Red.

– It's a red truck.

Can you make the truck roll on the table? – Roll on the table.

– Right, that's good.

What color are the tires? What color are the thetires Danny on the truck? – Wheel.

– The wheels.

What color are the wheels? – White.

– What else do you want to play with in there? Danny, Danny, here, what elsedo you want to play with? Wipe your nose.

What else do you wantto play with in there? Is that your favorite? Were you gonna give Big Bird a hug? – Give him a kiss.

– Give Big Bird a kiss.

That's good, that's good.

What color are Big Birds feet? Danny.

– Orange.

– Orange feet.

– Crawling.

– Your making BiG Birdcrawl, that's very good.

(Danny babbles)What? You're giving Big Bird a kiss.

Can you make Big bird drive the truck? – Drive a truck.

– Go ahead, show me how you can do that.

Make Big Bird drive a truck for me, okay.

Danny, can you make BigBird drive on the truck? Big Bird on the truck.

– Big Bird on the truck.

– Okay what truck is that?(Danny babbles) What color is that truck? It's a blue truck.

– Okay.

So, obviously you can't compare any parent to the other video, language-nutrition.

But what are a couple of thingsthat you noticed in this? Obviously this is afather who loves his kid, he's proud of his child's accomplishments.

He's actually the main parentin this particular case.

So what are some of the things you noticed in his interaction with his son? – [Man] He's having to do 90% of the work.

– He's doing a lot of the work isn't he? And that gets called, ifyou're coding in an experiment, that's what's calledintrusive or directive, right.

He's not following up on ways that he could really followup.

You know the kissingthe bird could have been an opportunity to goon, and on about kissing and loving, and hugging, and so forth but it was truncatedbecause he was directing.

Maybe if he was at home andhe didn't have our videos, you know camera running, he would have been alittle bit more responsive, but it's just to pointout how to directive.

Any other things that you want to notice.

– [Man] Gestures.

– Yeah, much more constrained, right.

Yeah, and he's proud and he's warm, but I think he's notspecific in his warmth is he? He's not, like, you'reshowing Big Bird love.

You're really hugging hard.

You know what I mean he'sjust kind of moving on.

So, I think these are some of the things that we would be looking for when we think about language nutrition for children with autism.

So in our scoping reviewwe asked three questions.

The first question wascompared to language nutrition in healthy and typicallydeveloping children, what's the languagenutrition like for children with autism and otherintellectual or delay conditions? Question number two.

What is the strength of the association between features of the language input and the language outcome? Is it as strong in this population? And the third, is to whatextent can we intervene to change or improveboth language nutrition, but more importantly, languageoutcomes or language health.

So, compared to language nutrition, comparing the length nutrition of typically developing children, and children with autism, here's a summary slide of what we found in the scoping review.

So we asked about quantity on the top, quality, and the qualityof the verbal interaction.

And interestingly, the literature says they're not that many differences in terms of adult word count, or in terms of thenumber of different words as an example of quality.

But what you see whenyou compare the autism to the typically developing children, is that the typically developing children get more questions, and the children withautism fewer questions.

Sentence complexity is higher for children who are typically developing and less complex for children with autism.

And like we just saw and wejust talked about in that video more direction towardthe children with autism, less direction more spontaneity and allowing the child to lead on the children who aretypically developing.

And just to show you, these different studies pick different indicators, so I just also want to show you that for children who have global delays or intellectual disability, similarly, not muchdifference in actual counts, or in the quality ofthe use of descriptions, but in encouragement, whichmight be in our interaction or in quality, more encouraging.

This may look a littlebit like a confusion, they're more encouraging forthe children with global delay, and that's because at the agewhen some of this was done, you shouldn't have toencourage kids so much.

The interaction is it's own encouragement, but here you have a lotlike, good Danny, good Danny, you know, over and over again.

And again less direction, and more direction.

Okay, so then what's thestrength of association between features of the language nutrition and language outcomes? And for this I'm gonna showyou a representative study.

This happened to be withchildren who had Fragile X, some of whom had autism, some of whom didn't but most had global developmental delay or intellectual disability.

And what you can see isthose parents who had high and sustained responsivity, had children whose rate of development wasfaster, and those adults who had low sustained responsivity, their children had slowerrates of development.

And when you look at our chart here, we saw that as sentencecomplexity went up, the child language scores went up.

When commenting on what the child was paying attention to went up, the child language went up.

When redirecting thechild's attention was high, the child outcome was lowwith a negative association, and I want you to payattention to the last one.

When parents use what's beencalled telegraphic speech, that's the, I want cookie.

The outcomes were worse for children.

So the positive relations withthe child language outcomes are seen for very long periods of time after the initial observations were made.

So now the question is towhat extent can we change or improve language nutrition, and thereby improve child outcomes, 'cause it could be that those associations have something to do with the child.

The child pulls from the parent.

So, for this I want toshow you two meta-analysis.

So this one on the left, remember on these force plots that you put the difference betweenthe target population, in this case children withintellectual disability and autism are grouped together.

The difference between those children, and typically developing children is the symbol in the middle, The lines emerging from that symbol.

are the confidence intervals.

And if there's a positive difference the lines stay to the rightof the red line, the zero.

And if there's a negative influence, the lines are to the left.

So you can see when youlook study by study, there's not a lot of studiesthat show big change.

But when you add in a meta-analysis across all those studiesyou get a slight advantage.

The blue diamond, whichis on the positive side.

And here's another study.

This is just with children with autism.

And you see, they had itarranged a little bit differently in their paper but it'sthe same general idea, and you see again a little blue diamond, that's to the positive side, but many studies don't show an advantage.

So, why the modest effect sizes? Why aren't we seeing more? So one thing is thatin all of these studies children are enrolled in other services.

So you have to havethe language nutrition, really make a contributionabove everything else that the children are getting.

So it's a tall order.

In many of these studies languagenutrition is very minimal so there's very few sessions, or the parent education takes place without the child being present.

And there's limited modeling, and limited feedback about how the parents are doing.

And I think another reallyimportant thing here is that a lot of the studies only choose one element of language nutrition and many of them choose the quality of the verbalinteraction, which is great.

We want a fun verbalinteraction like we saw in that first video, but if you don't give childrenmeaningful language to hear they have nothing to growtheir own language from.

So I think we need tothink about interventions that have all elements oflanguage nutrition represented, and another point is thatwe should be looking at what we think we can change.

Which is not primary autism symptoms, because that's not whatwe're intervening on, but rather the language outcome.

So I'm actually beginning to pull together some ideas for a randomized trial where we can see whetherif we do language nutrition in a really meaningful way, whether we can improvethe language outcomes.

And then we would have stronger evidence about how to work with parents.

And the ingredients that Ithink are gonna be important, are that when we talk withparents about language nutrition, we have the caregiversand the children together.

That way we can model what wemean by language nutrition.

We can get going a funny, fun filled, responsive interaction.

And we should make it as explicit as we can for parents and their child and sustain it over time.

And one of the thingsthat I'm hoping to do is use that LENA device to get some ideas about how parents are using language so then we can personalizethe language nutrition for an individual family.

But I think there's enoughevidence in that literature that we can begin talkingwith parents right now about how they can improvethe language nutrition for their children.

So here's some general suggestions.

One, parents should bewarm and responsive, have fun, like father number one.

Provide the child with manywords, and many gestures.

You know in the earlyphases of language learning and the typically developing child, they might hear a word3, 000 or 10, 000 times before they learn car, or shoe, or dog.

So let's give children withautism that same opportunity.

Repeat the words frequently.

Follow the child's focus of attention, rather than direct the interaction.

Use full sentences, useappropriate grammar.

They don't have to be complex sentences with really complicated vocabulary, but at least they should beproper English sentences, so the child learnsproper English sentences.

Avoid excessively simplified speech, like the telegraphic speech.

Avoid directing, andespecially avoid prompting.

I think when parents get used to prompting it's very hard for them to fade.

Model using many grammatical sentences, and using the words overand over in a single time.

And think more about whetherthe child's understanding, rather than whether thechild can say the word.

I think that if you're working with people at different stages ofdevelopment you might think about at the early stages, which in typically developingchildren is years of life but maybe in childrenwith autism is shifted toward older years, but still early stages.

Think about lots of words, and warmth and responsiveness, and maybe some gestures.

And then as children get alittle bit older or more mature, then add in a wide variety of words, longer and more interestinggrammatical sentences, more questions, and talkabout past and future.

Thank you all for your kind attention.

(audience claps).