Interview with Dr. Walter Zahorodny

SafeMinds recently had the privilege of speaking with Dr. Walter Zahorodny, Principal Investigator of the New Jersey Autism Developmental Disabilities Monitoring (ADDM) Network about his latest research, which assisted in calculating the newest CDC autism rate. As highly publicized last March, 1 in 54 or almost two percent of all American eight-year-olds now hold an autism diagnosis. To shed more light on this situation, Dr. Zahorodny shared his incredibly important viewpoints surrounding the ever-increasing autism crisis with SafeMinds. To learn more, watch our newest video which is an excerpt from a 70-minute interview. The following is a transcript of that interview.

Rebecca Estepp

Okay. All right. Hello, this is Becky Estepp from SafeMinds. I am here today with a special audio and visual version of SafeMind Shares. We hope you all find this new format valuable. The CDC just released their current autism rates from the autism and developmental disabilities monitoring network, which I will now refer to as ADDM. The new rate is staggering, one in 54, almost 2% of American eight-year-olds have an autism diagnosis.

This new rate represents a 10% increase from the prior reporting period announced just two years ago. And sadly since ADDM has been tracking autism rates, which is 20 years, the rate has increased 175%, or in other words, there is now 2.76 times as much autism as there was in the year 2000. Boys still outnumber girls by approximately four to one, which means that 3% of American boys age eight are affected by autism. We at SafeMinds want to know just what’s going on here, so we’re going to talk to an expert. Our guest today is Dr. Walter Zahorodny, principal investigator of the New Jersey autism study and associate professor of pediatrics at Rutgers New Jersey medical school. He is a clinical psychologist with 25 years experience working with children and families. Dr. Zahorodny main areas of research include autism surveillance, the development of autism screening tools and identification of risk factors for autism. Welcome Dr. Zahorodny, and thank you for talking to us today.

Walter Zahorodny

Hi Becky. It’s a pleasure to join you.

Rebecca Estepp

Okay. Just to make things clear for our listeners, your name is on this latest ADDM report, because you’re in charge of reporting New Jersey’s numbers. Do I have that correct?

Walter Zahorodny

Yeah, that’s right. I’ve been involved with this New Jersey surveillance team from the beginning, from 2000. And I’m currently the PI [Principal Investigator].

Rebecca Estepp

Okay. We’re so happy to have you here today because that holds a lot of gravitas. So I just feel honored to be getting your insights. So let’s get to the questions SafeMinds has for you. And number one is how is the ADDM data collected? I’m not sure if everyone in our audience knows.

Walter Zahorodny

Yeah, well all the States in the ADDM network, including New Jersey, they use a common method for case finding. The way we do it is by working with one age cohort of children at a time. And in each surveillance region, in our case, it’s four counties in New Jersey. We review all the possible health and education records of children who were born in the right year.

And a group of researchers goes to these hospitals, developmental centers and schools in our region and they wind up reviewing the records of not only children who’ve been diagnosed or classified with autism, but in the case of the education system, children with any special ed classification. And with regard to the health developmental centers, we look at the records of children who have received any developmental, neurological, behavioral or psychiatric care. And it’s a two phase process, once our researchers review the records, they identify the records, the children who have some indication of possible autism and they copy out all the information that we can find in those records that could be useful to understanding whether the child has autism and what the characteristics of the child are.

And we put all of this information into a chronological file and we strip out all the identifiers. That is information that could tell people what town the children are from, who the doctors are, teachers are. And each child’s portfolio of information goes to a special panel of experts, this is the phase two of the two phase process. And those experts use CDC designed procedures to identify which ones are true cases of autism and which ones are not. And they score a huge array of information so that we can know the characteristics of these individuals with autism. And so we go through our whole region and ultimately we identify how many children meet our definition of autism and then we divide that number by the total number of children in that region, and we’re usually focusing on eight- year-olds. So the total number of eight-year-olds would be the denominator and the identified cases are the numerator. And that’s the process for determining the prevalence rate.

Rebecca Estepp

Wow. There’s a lot that goes into it and that’s interesting. I was going to ask you about the anonymity of that. So the children’s names are stripped out in there so they’re not identified.

Walter Zahorodny

Right. The children’s names are out. Anything related to families, names of siblings. We also take out any information that would identify the race or ethnicity of the child so that people would not have that other information and possibly biasing their scoring or determination of cases. And in a place like New Jersey, there’s a lot of information, both from the school systems and from the clinical providers. I do everything I can for confidentiality, because, of course, we could only do this if we had the trust of our sources and the sources would only be likely to participate with us if they knew that we would keep all of that information that needs to be private, confidential.

Rebecca Estepp

Okay, great. We had a little bit of… Your cell phone got, I don’t know, like the signal wasn’t that great. But I think we got most of it. So do you think you’re in this strongest cell reception by in your house, do you know? I Kind of think so. I’m not positive.

Okay. No problem. Okay. So my next question is what do you think the strengths and the weaknesses are of the ADDM methodology?

Walter Zahorodny

The strengths are, because of the way the study is designed, we’re able to identify not only children who have been diagnosed with autism, but ones who satisfy the case definition but don’t have an autism diagnosis. Another strength is that we’re able to recruit a huge amount of case specific information about the characteristics of these children. We know for example, about their diagnostic history, when they were evaluated, what professionals evaluated them, what characteristics of the individual are in terms of are there other conditions or props in addition to autism?

And so we form a very comprehensive picture of each child and we’re able to identify children, as I said, who haven’t been diagnosed. So I think our ability to be accurate is very high because of those two factors. On the other hand, it takes us a very long time to gather this information, because in New Jersey, we find a huge array of professional evaluation and school-based studies and it just takes a lot of time to go through that and to assemble it and analyze it. So I would say our weakness is that it may take a long time to get to the end point, which is often a report on prevalence.

Rebecca Estepp

Got it. So what kind of changes have you seen in the ADDM approach over the last 20 years you’ve been an investigator?

Walter Zahorodny

Well maybe this is another strength that we’ve been able to use the same methodology with the same case definition of autism all the way through the process. And that gives us confidence that we’re using the same strategies and procedures to find the cases and that makes our data comparable and understandable from year to year. However, the amount of data we are reviewing and analyzing has grown a lot. So that has placed a bigger burden on the network, all the researchers at every site to do the work that’s necessary. I would say that the amount of information that we process now is maybe five to six times greater than what it was in 2000 when we started.

So this is not to take anything away or to add anything, but it’s become more and more difficult, I would say, to do surveillance in our region as comprehensively as we would like and still make our projected deadlines for completion. At one point, the network experimented with an approach that would maybe shorten the amount of time necessary. That was called streamlining and that experiment left some things to be desired. So we went back to using the conventional system and we just try to do as best we can in the time we have. We haven’t really had any significant alterations in the comprehensiveness of what we’re doing or the primary procedures that we use.

Rebecca Estepp

Okay. And are there any changes in the ADDM methodology coming up in the future?

Walter Zahorodny

Yeah, very big methodological change is coming. The reports that were issued today may be the last reports, probably will be the last reports in which the ADDM network uses this active case finding strategy, which identifies cases of autism that haven’t been diagnosed. Starting with the next cycle, starting with 2018, we’re actually going to be doing a case that’s only based on counting children who have already been diagnosed.

Rebecca Estepp

And you just said 2018, and I just want to make sure everybody understands out there that means 2018 data of eight year olds.

Walter Zahorodny

Yeah, that’s right. So our next cycle is a study year 2018. And that will be based on children who were born in 2010 and were eight year olds in 2018. At the very beginning of this cycle which started in January, the CDC told the cooperating investigators in the network that we were not to use the old procedures. That we would be provided with a new definition and the new definition is limited to children who’ve already been diagnosed. And we’re also not going to be comprehensive in the way we recruited all that wealth of data.

The amount of information about other issues or the presentation of autism will be dramatically limited. And the CDC is shifting to this new methodology and new definition because they say this will help make us be able to report in a more timely way. Takes us a long time as I said, to go through all of that data and to do so carefully. The new methodology also will be a single phase process. It won’t involve the evaluation of information by experts anymore. Pretty much the researchers working in the field will be making a determination as to who’s a case and who’s not a case immediately when they’re recording information. And that may be a limitation of the new approach that is it will be possibly circular for individuals to do both the reviewing and collecting of data and also making the case determination of who’s a true case and not. I think the old approach, the traditional two phase approach had a built in safeguard. It’s that the experts ultimately were determining which children are cases and not cases.

Rebecca Estepp

Interesting. Well we’ll have to see two years from now what those changes are going to be like. So that’s very interesting. I’m a parent of a child with autism and he is 22 years old. So pretty much I’ve been looking at the work that you’ve been doing for 20 years. Some parents get a little bit confused because the CDC releases autism rates in two different ways. Because there’s the National Survey Childhood Health rate and that rate is usually a little bit higher. That one says one in 36 children have autism, and then there’s also the ADDM rates. Can you discuss the difference between those two because it really, I think, perplexes a lot of people out in the community.

Walter Zahorodny

Yeah. I’m a little perplexed myself. I’m not surprised. So those National Surveys – they develop their information on the basis of calling parents or asking parents whether they have a child at home with autism or some other disability. There is no verification of that information and there is no collection of case information. What are the individual child’s characteristics and how severe? Are there other co-occurring conditions? Those studies, those surveys, really, are going on a parent report at a single period of time. I’m pretty sure that they’re not really considered epidemiologic surveillance. These are more like polling data rather than hard evidence-based studies. And so I’m a little bit confused why… So we track very consistently, not only the individuals who have autism, but we know which ones have been diagnosed with autism. And at no point really does the data from the national surveys correspond with our information at a state level or at the federal level.

On a federal level. I’m not an expert on who is called, how they form their information base, but it seems incongruous to me that a parent survey would yield higher prevalence than an approach that depends on active case finding. I don’t want to disqualify the NHIS or the NSCH reports, but to me they’re not really accurate or dependable for epidemiologic information or for public planning information.

Rebecca Estepp

Okay. That’s really good information to have. So thank you for it.

Walter Zahorodny

Your point is very well taken. Why is it that we have two or three different estimates and they are always in disagreement. Sometimes the disagreement is diminished and sometimes it’s accentuated. But I also can’t understand why a parent report survey would generate higher estimates than an estimate based on active case finding.

Rebecca Estepp

Yeah, it’s always made me kind of scratch my head too, other than I don’t know why a parent would misreport or misrepresent.

Walter Zahorodny

Mm-hmm (affirmative).

Rebecca Estepp

I often think about that, but I do kind of usually feel a little more comfortable with the ADDM rates.
Walter Zahorodny

That kind of makes me feel like the work is worth it.

Rebecca Estepp

Yeah. Okay, back to the ADDM Report. So prevalence for this report varied by two and a half fold by site. Colorado reported a 1.3% rate while your site, New Jersey, reported 3%. For me, a mother of a boy, that means 5% of boys in New Jersey have a diagnosis. So can you give us any guidance? Are there differences in the way data is collected in different states?

Walter Zahorodny

Yeah, that’s a good question. Initially it was the plan of the CDC to require all of the cooperating sites to have access to and to collect information and analyze information from both health sources and education sources. That way, there would be the most likelihood of a robust, reliable finding.

Some states in the network, like New Jersey, have access to both health information and education information. In some states, the investigators don’t have permission to review education records; and so on the basis of that single difference, there will be big differences in the number of cases that can be identified.

Another difference stems from the fact that there could be, and I’m sure are, large differences from state to state as to what’s considered normal or expected or acceptable information for health or education evaluation. Reports can be full of detail and well documented and give a very complete picture of the child or the individual being reported on. In other places, the information is more scant or limited.

The methodology of the ADDM Network is definitely affected by both having access to the greatest amount of information and to the quality of the information that’s found.

I’m certain that New Jersey, obviously we have access to all the education and health sources, so that’s a big advantage. But I think quality wise, our researchers, I’m sure, are finding both more information per child and also more complex and robust information from different specialists.

For example, in New Jersey, a child undergoing an evaluation for special education services is evaluated by a child study team made up of professionals from different backgrounds. A speech pathologist will provide an evaluation as well as a psychologist, a physical therapist, a social worker… And each of those professionals brings their own expertise and their own interests and experience to the picture. By virtue of that, we get very complete, well-detailed pictures, elaborated descriptions of children; whereas that’s likely not to be true in many of the other states.

In my opinion, it can’t be really that the true rate in all of these states is so divergent from other rates. It’s a reflection of the fact that there are differences in data quality and access to information.

Rebecca Estepp

That’s what it looks like to me. And in my opinion, I would think New Jersey, you all seem to kind of be like the gold standard and it’s not… I would imagine that the New Jersey rate is actually more representative of the entire country, is what I’m trying to say.

Walter Zahorodny

Yeah, you’re right.

Rebecca Estepp

Yeah.

Walter Zahorodny

You’re right. I think we’re probably closer to what the reality of the situation is. We’re really generating estimates. There is no exact number, which is exactly correct, but we’re probably getting closer than any other state to estimating accurately because we have better quality information.

It’s not that New Jersey is more risky or dangerous with regard to autism… It’s that it’s more possible for the researchers to identify the true cases given the quality of the information. It kind of stands to reason that a state that spends twice as much as another state on education will generate more qualified evaluations, more correct information, and also yield more right diagnoses.

Rebecca Estepp

I agree. So when I was reading the ADDM report, they talk about the different states. There’s 11 different states. But they also use the term site. Can you kind of explain the difference between a state and a site?

Walter Zahorodny

Yeah. Well, only one state in our ADDM Network, one only one group of researchers in our network covers their whole state… That would be Arkansas. The researchers in Arkansas cover the whole state. Every other site only covers part of a state. For example, New Jersey has 21 counties, but we do surveillance in four counties.

Our site is in New Jersey, so sometimes I use the words state and site interchangeably; but in reality, every site except Arkansas is only surveying part of the state. The reason Arkansas is doing the whole state is because according to our original agreement with the CDC, the CDC stipulated that they wanted sites to join the network who has a birth cohort of at least 25,000 births a year; and to get to the 25,000 births a year, the researchers in Arkansas needed to cover the whole state because I guess that’s close to their annual birth population.

New Jersey is a large densely-populated state with a much higher population, so we wouldn’t be able to cover the whole state. So we’re covering what’s essentially a quarter of the state population by covering four counties.

Rebecca Estepp

Okay, thank you. That clears up that situation perfectly.

Walter Zahorodny

Another thing I can add is that I think it’s a big advantage that in New Jersey we’ve maintained the same region for surveillance. We’ve surveyed the four counties each time, even in the most recent report. Some of the other states, however, have shifted their focus or their targeted surveillance.

Even, for example, Georgia. The CDC runs the surveillance in Georgia. They used to cover five counties of metropolitan Atlanta. In the last cycle, they dropped down to covering only two counties. So those kinds of shifts in what’s the population of the region complicates things in terms of understanding whether there’s real change or not.

We’ve tried very hard in New Jersey to maintain not only the methods and procedures where we’ve been successful in maintaining our surveillance in four counties.

Rebecca Estepp

Great. So I wanted to know how are the states selected and are they representative of the US population?

Walter Zahorodny

Yes. From the beginning, the members of the network were selected on the basis of applications or proposals to do research from different states. Only states that wanted to do surveillance participated in the competition to be granted an award.

Many, I would say most states, probably didn’t compete for whatever reason, for whatever combination of reasons. But there were states and the CDC selected the states that would be funded to do surveillance in the same way that independent researchers compete to get a grant from say a federal agency or a philanthropy group. That is, they send the proposals to a panel of experts who then say, “Okay, we’re going to rank them in the quality of their presentation or the quality of their proposal, and then the CDC makes that the basis for their funding.

So if there were 20 states who submitted proposals to do surveillance and the CDC had funding sufficient for 10 sites, they would pick the 10 sites that had the best scores from some panel that they established of independent experts.

Rebecca Estepp

Okay.

Walter Zahorodny

And it’s definitely not representative.

Rebecca Estepp

Okay.

Walter Zahorodny

As you can quickly tell, there’s a predominance of Southern states in the network and the 11 states that are in the network are not geographically representative or representative of the whole United States.

Rebecca Estepp

So what are the implications of it not being representative of the United States?

Walter Zahorodny

It’s hard to say that the overall findings are generalizable. If the states were selected with an eye toward being representative, then the overall estimate would be closer… Could be believable as representing the risk of autism across the United States. As it is, what the network can say is that these are representative rates for those particular sites or states and we’re just taking a mean across them when we cite the overall prevalence estimate.

Rebecca Estepp

Got it. So a colleague of yours, Dr. Cynthia Nevison, had a study released last week and she used the California Department of Developmental Service (DDS) data for her latest study examining autism rates in different socioeconomic and racial populations. DDS data comes from actual cases filed with the state. Do you feel that DDS data out of California might be more robust than perhaps ADDM or the NSCH?

Walter Zahorodny

I think that the California data, the CDDS has a couple of strengths, but I don’t think it’s necessarily representative of the true number of autistic individuals in California. The overall estimate is quite a bit lower than what you would expect given what we see in New Jersey.

I know in part that’s because the California set, the CDDS, really registers or tracks children who have significant impairment due to ASD. Perhaps the California set isn’t complete in that it doesn’t include individuals who are more mildly affected.

I’m not really sure to what extent the CDDS is representative of California on a socioeconomic level. My suspicion is that it probably represents low and moderate income families pretty well but may not represent autism rates in higher SES families.

So those are some of the limitations of that set. On the other hand, the CDDS database is more up to date; that is, it comes out every year and I think it functions as a pretty good early indicator of where autism trends are going because they report on younger children and they report every year.

I kind of like that, and I think that one can see things about trends possibly sooner by looking at the California data then looking at the ADDM data, but I’m pretty sure that the California data is understating the true scope of autism by maybe 50%. I think if you were to establish a prevalence estimate based on California data, it will be closer to 1.5% than the 3%.

Rebecca Estepp

Got it. Okay. According to this report, what was the average age of diagnosis?

Walter Zahorodny

So in our 2016 ADDM findings, the average age; that is, the median age of autism diagnosis was 51 months. That’s really not different than in previous cycles. I am pretty sure that going back at least 10 years, or maybe even further, the median age of autism diagnosis was in the 51-52 month range. So in my opinion, there’s really been no change or no improvement in the age of first diagnosis.

Rebecca Estepp

Okay.

Walter Zahorodny

Kind of a surprise because if you assert that there is a lot of better awareness being brought into the picture of determining who has autism, you would expect markers like age of autism diagnosis or age of first evaluation to be improving over time, but that’s not something we really see.

Rebecca Estepp

Right. That’s too bad that it’s not getting better because basically the earlier you get this, the more time the child has for therapies and intervention and possibly to better the outcome.

Walter Zahorodny

Right, definitely the earlier the better for sure.

Rebecca Estepp

Yeah. And I also read that this report used the DSM-5 definition of autism while the other reports use the DSM-4, could that have impacted the rates?

Walter Zahorodny

Well, we did use the DSM-4. We used both definitions actually in this report and in the preceding one, the one from 2014. And in the report from 2014 there was a slightly bigger difference. There was a more meaningful difference between the estimate by DSM-4 versus by DSM-5, that being that the DSM-4 was more inclusive or yielded a larger estimate than the DSM-5. But in the more recent cycle, that is in 2016, I see the disparity between the DSM-4 count and the DSM-5 count shrinking. So I think it’s the shift to the DSM-5, I would say it isn’t having a dramatic or a significant effect on the estimate and possibly, as we go forward, if we were to use the same process, we would probably see the difference shrink even more.

Rebecca Estepp

Okay. That was a little surprising to me. I thought that it might’ve had a greater impact, but it doesn’t sound like it did.

Walter Zahorodny

No, not really. I guess I would tell you though, it’s largely because the DSM-5 definition encompasses automatically all children with a previous autism diagnosis. So how the number and pattern of the definition has changed, but as an aspect of the DSM-5 definition, they automatically count every child who had a DSM-4 diagnosis. So they’re kind of locked together in a way that makes it a little bit tricky to determine is it really the definition that’s different or some other factors.

I do think that the DSM-5 definition is more restrictive than the DSM-4 definition, and I see that more in the four-year old cohorts than in the eight-year-old cohorts because obviously fewer children at age four will be diagnosed. So the DSM-5 criteria are too narrow or too restrictive, I think, for helping us identify young children with autism.

Rebecca Estepp

Actually, can you tell us a little about … There was a companion report on four-year-olds, and that was the first time I ever saw that, can you give us a little idea of what that was all about or why that occurred?

Walter Zahorodny

Yeah. So there was a second report revealed or released today using the same methodology in a subset of the ADDM states. Not all 11 participated. It was a group, I think, limited to six states, six sites in the network, and they used the same method and procedures, and they report … So why do they do that? At age four, you know you’re not going to be able to find all the individuals with autism because they’re not all coming to attention before the age of four. Average age of first attention is still somewhere in the range of four years, so by definition you’re not going to be seeing maybe half of the children with autism if you’re looking at that age.

But the advantage of looking at a younger cohort like that is, I guess, two fold. One, it helps you learn more about the early process of evaluating and diagnosing autism and it also possibly gives you a window on what the future estimates will be for eight-year-olds. There have been, interestingly enough, not that many people are aware of this, but this is really the fourth time that the ADDM network has also looked at four-year-olds.

There was a report published in 2019 in which they looked at four-year-olds at age … Sorry, four-year-olds in 2010, 2012, and 2014 and yet today there’s this fourth approach or this fourth attempt to look at autism in this younger cohort. And this one also tells us that autism is increasing and it tells us a little bit more about what the characteristics are of the children who get identified at this age. As you would expect, children with autism plus intellectual disability tend to be identified earlier and so the four-year-old cohort is likely to be more impaired or to have multiple problems, but definitely not equivalent to a complete count or a complete estimate, which is more likely to be determined when the child is eight years old. What the four-year-olds allow us to see is the trend going to plateau or go down or go up.

And the most recent report, today’s report, gives us clear indication that this is not plateauing or going down rather that, it’s very likely that autism is continuing to go up. And when we get to this group, when they’re eight-year-olds, the rate is going to be obviously quite a bit higher than when we looked at age four.

So in New Jersey for example, the prevalent estimate for this four-year-old cohort was 2 1/2%. That’s indicating to me that when this cohort is surveyed more completely at age eight, the rate will be somewhere above 4%.

Rebecca Estepp

Wow. And my next question does have to do with intellectual disability, which you just touched upon. So in this ADDM report, it showed that autism is still four times more prevalent in boys than girls, but this time we saw that girls on the spectrum are more likely to have intellectual disability. It was 39% for girls versus 32% for boys. Is this the first time we’ve seen this?

Walter Zahorodny

I would say no, however I can’t point to the report in which this was made, but I’m pretty certain that we observed this earlier in the network, that is that girls are more likely to have intellectual disability compared to boys with autism. I don’t think this is the first time it’s been revealed, but maybe this is the first time it’s getting appropriate attention. I wish I had the ability to tell you whether we found this in 2010 or 2012 or possibly earlier, but this difference in autism in girls, we kind of suspected it and knew it before the publication today.

Rebecca Estepp

Okay. So this time there was no difference in autism prevalence between African American and Caucasian children, can you tell us a little more about this and if there were other racial differences that emerged from this data?

Walter Zahorodny

Yes. So for the first time, correct, it is correct that for the first time, no state in the network … Or sorry, the first time across the network, it was clear that boys … Sorry, that children, white children and black children who were equally likely to have autism. But we’ve seen that earlier in New Jersey. I think as early as 2008. We saw that is the report in 2008 already showed in New Jersey that the rate among minorities was equivalent to that of white children. So New Jersey was an early indicator of that. And now the ADDM network, as a whole, is seeing that there is no likely difference in rate by race.

Walter Zahorodny

However, in most of the network, even now, there is a disparity such that Hispanic children with autism are less likely to be identified. In New Jersey, we find that Hispanic children have the same prevalence as white and black children, but in just about every other state, there’s a significant difference and a lagging difference, which suggests that this is an important area that we call for public health attention. There is something that needs to be done or should be done to identify the Hispanic children with autism, which were often invisible in their states even in 2016.

Rebecca Estepp

Wow. Okay. Were there any other differences in this ADDM report than previous reports that really stuck out to you?

Walter Zahorodny

Well, the most startling aspect of the report was that in New Jersey … I see the New Jersey rates as being the leading indicator and I was surprised to see that even though our rate overall didn’t go up very dramatically between 2014 and 2016, our rate of autism in boys was at the 5% level. One out of 20 boys have autism. That’s hard to get your head around. Sometimes I find myself checking myself and questioning, why am I saying that? Well, I’m saying that because the data are clear that that’s the case. I certainly never thought that I would be identifying such a high rate of autism among children in our state, but that’s certainly the case. I think that would be, from my perspective, the two big findings of this report are that Hispanic children are significantly under-detected in most states and that autism has the leading indicator shows that one out of 20 boys has autism.

Rebecca Estepp

It is, it is shocking. And right now, because of COVID, we’re hearing about flattening the curve, and it’s difficult as a parent of an individual with autism and watching our rate climb, climb, climb, and to be hearing the country talking about flattening curves, and which absolutely everybody wants to happen with COVID, but it’s hard. It’s hard when our rate has been climbing and it doesn’t seem to really get the attention that it deserves. I’m sure you probably feel similarly.

Walter Zahorodny

Yeah. I guess one of my dominant emotions is frustration because I really have a hard time accepting or understanding why a dramatic escalation in autism prevalence isn’t experienced as an important or urgent public health issue. How can we say that 3% of the population, 3% of children have autism or 5% of boys have autism and not treat it like an emergency? It’s not a trivial phenomenon. It’s not like we’re talking about individuals who are just somewhat different or have different learning strategies or different temperaments. These children have very significant learning problems and issues related to social involvement and capacity. I just don’t understand why it was shocking for people to realize that 0.6% high autism in Brick Township was an urgent problem, but three times higher rates across the network and 3% in New Jersey isn’t distressing or shocking. If we said that 3% of the children in our country had a hearing impairment or a visual impairment, I think that would be taken to be a real urgent issue that would call for lots of research and lots of investigation and intervention or some complex combination of reasons, I don’t know what to call it, the meme of it’s better awareness seems to have taken hold and the media repeats it over and over, the detriment of understanding the reality of the situation. I.. [crosstalk] propaganda, but the CDC, I think, has been just understating the issue, but really it’s the behavior of the press that seems unconscionable to me and that they take for granted that a claim or the assertion of better awareness explains the problem is sufficient. It’s really not a sufficient explanation.

Rebecca Estepp

Well, when the principal investigator is saying that it’s not just better awareness or better ways of diagnosing, you would think that that should start such a discussion in this country and yet it doesn’t, and it’s astounding. So in comparison, how does the rate of autism compare with other childhood illnesses or conditions? Do you have any insight into that?

Walter Zahorodny

I guess I would really only be able to reference what’s in the public domain. And of course, there are overall papers from people at the CDC on the prevalence of other childhood disorders and it’s really only autism that shows the pattern of rapid escalation. ADHD is a prevalent phenomenon problem, a prevalent disability among children possibly, I don’t know, in the range of 7% or something and it has increased, but the increase has been, I would say slight or moderate. Whereas the increase in autism is dramatic. Intellectual disability really hasn’t increased or changed significantly. That is the proportion of children with that issue. CP, cerebral palsy. I read a paper recently which seems pretty convincing that the prevalence of CP in the United States is decreasing. Other impairments, visual impairment, hearing impairment, those are static, flat over time. No increase or decrease in those things. Autism is unique in the sense that its growth has been so dramatic and so broad.

Rebecca Estepp

Yeah. Do you feel our country understands the scope of the autism problem?

Walter Zahorodny

I don’t think so. I think the campaign to minimize the escalation in autism prevalence is successful. I think the persons in our communities, the individuals who don’t have family members or children affected by autism hear the escalation is a matter of better awareness and that’s what they accept. When our initial findings came out in 2007, that would be the findings from the 2000 and 2002 study year. Those findings came out in February and by July or August the politicians in New Jersey had assembled a portfolio of seven autism laws. Laws that made autism a reportable condition. A law that made for funding autism research and other initiatives passed unanimously within a couple of weeks and were in law, made law, the same year that our prevalence estimate of 1% was public. At 2% nobody cared. No politicians were contacting us about what changes were needed and I’m pretty certain mainly because of coronavirus that the findings of 3% are going to go without action.

Rebecca Estepp

That’s awful.

Walter Zahorodny

Yeah.

Rebecca Estepp

So I heard you say autism should be considered an urgent public health concern.

Walter Zahorodny

Right.

Rebecca Estepp

I’ve always wondered why autism is never… the term epidemic is never used for autism. Do you feel it’s an epidemic or do you use public health concern for a different reason like, the label?

Walter Zahorodny

Yeah. So epidemic, the term, the word has now a very strange political connotation. I do think it’s an epidemic in that it’s an increase in prevalence for unknown reasons. We definitely see a very dramatic increase in prevalence and we certainly don’t understand the reasons, so I have no problems saying it’s an epidemic. Sometimes the term is overused, so I try to shy away from a trivialization. I like the term surge or wave because those haven’t been used as much, but they all refer to the same phenomenon. The CDC seems to have an aversion to calling autism an epidemic. They’ve called other problems, including a prevalence of sexually transmitted disease, obesity, childhood obesity, and other problems, they’ve had no difficulty… Drug use, opiate abuse, they have no problem describing those as epidemics.

But for some reason probably the leadership of the CDC can’t accept that term in public. I suspect it has a lot to do with the fact that they see their primary purpose as being mass vaccination in the United States, and phenomena which seem to impair or interfere with mass vaccination are unacceptable. So they probably shy away because of the battle over the question of was the autism epidemic caused by some vaccine related issue? I think they never got over that. And probably there’s some directive they have that went down to people saying never use that term in public when it comes to autism.

Rebecca Estepp

It would certainly seem like there’s some sort of directive because in the 20 years since my son was diagnosed and in the 20 years that you’ve been doing your research that they stay away from that term when I think it’s so urgently needed. So thank you for that answer.

Walter Zahorodny

One could do an interesting study or an essay on just the way they’ve used that term. I’m pretty sure in the first two reports it was considered an urgent public health concern or urgent public health problem. And then later autism was I guess reduced to an important public health problem or concern. And in today’s report, in the eight year old report… Let me see, I think they manage, sorry, should have had this in front of me… it’s no longer urgent or important. It’s something like it’s a public health issue but no longer reported urgent.

Rebecca Estepp

Oh my God.

Walter Zahorodny

So to me that conveys… Somehow that conveys to me the sense that the CDC either doesn’t see escalating autism as a problem or they want to diminish the importance of the phenomenon somehow. They have no choice, I guess, but to record it because it is increasing in all of the states and increasing by every approach that tries to look at time trends. But for some reason autism can’t be, I guess, important enough to call for that urgency or for importance, because when something is urgent or important you try to do more to solve the problem, to discover the reasons for it and to do something to change it. I see their public health statements are very weak on the point of autism and autism prevalence.

Rebecca Estepp

Yeah. It’s no wonder so many families dealing with autism have lost faith in the CDC when they’ve asked for help and we just don’t get the attention that you would think would be deserved so.

Walter Zahorodny

Right. They say, and to some extent they have a right to defend themselves, they said, well, we have been making a lot of public information available. They have the “Know the Signs. Act Early.” program. They’ve put a lot of resources into that over 15 years. I don’t know whether it’s effective or successful or not, but they certainly spent money on it. And they also conducted, I think, very expensive research into perinatal risks for autism under the aegis of the SEED or CADRE program, which I think yielded nothing substantive. So they did spend money, but it was probably not so consequential and definitely did nothing to stem the increase of autism.

Rebecca Estepp

Right. Okay. So back to the soaring rate of autism and how it’s risen so quickly in such a short amount of time. Doesn’t that point to environmental factors being implicated in this increase?

Walter Zahorodny

Yes, in my opinion, definitely. Environmental factors are definitely implicated or likely in this situation, very unlikely that genetic factors would change radically or dramatically. Genetic changes occur over generations, not over a 20 year period. And even if we were to look only at since 2010, the rate has gone up so significantly. Must be because of some change in environmental risk or triggers. I wish I understood what those risks and triggers were or are, but it seems like they’re in play across the United States and they’re consistent. There’s never been a decrease, never been a decrease in autism prevalence estimates and…
[disruption and rephrase of question]

Rebecca Estepp

Okay. With the rate that has risen so quickly and in such a short amount of time, do you feel that environmental factors have to play into the increased rate of autism?

Walter Zahorodny

Yes, definitely environmental factors are implicated in this increase of autism. Genetic factors would be evidenced or active possibly over generations, not over a 20 year period. And even if we restrict ourselves to, say, looking at autism since 2010, it’s clear that it’s increased significantly and broadly across the United States, across all ethnic and racial groups. And this can only happen if there is some important triggers or risk factors for autism increasing very, very broadly. And the increase is cumulative. That is, it never goes up and down. It seems to be building only on top of the previous rates. So I don’t know what environmental factor or factors are involved, but definitely its effects are very significant.

Rebecca Estepp

Excellent. And this is your last question. What do you think needs to be done to make this situation better?

Walter Zahorodny

Well, since I’m a clinical psychologist as well as a researcher, I would say first of all, we should have and promote universal autism screening at 18, 24 and 36 months. There is no better way to overcome the disparity that we see based on race. Then what would happen if we were to do universal screening at an early age with a good screener? That would help us a lot and that would be the step that would mean the most in terms of getting the most children into interventions.

The other thing I would recommend is, since we haven’t yet seen the plateau of autism prevalence, I would want to continue at least in some way the active case finding strategy, which allows us to see also those children who haven’t been diagnosed with autism. There’s 20 to 25% in every cycle of children who have autism, by our surveillance activities, who haven’t been diagnosed. We need to… If you really want to know the scope, you have to go beyond counting the diagnosed cases.

Another thing I would like to do is to establish the capacity or more intensive research into environmental risk factors and triggers for autism. People have started that, maybe they’ve exploited the low hanging fruit, but there is something to discover by some unconventional, maybe research or I don’t know, a novel research into possible risks and triggers.

And finally I would encourage that we try to follow these cohorts, these large cohorts of identified children with autism over time. A longitudinal study which would hopefully tell us something about what are the successful interventions or the treatments or the constellation of individual characteristics with treatments that are most effective. Because all of the children we identify at age four or eight are going to likely stay impaired by autism in later childhood, adolescence and adulthood. And we need to start learning about what happened to those and what interventions and changes lead to better outcomes.

Rebecca Estepp

Absolutely. Well, Dr. Zahorodny, thank you for this very thorough interview. Your answers were superb. I know that people listening are going to gain a lot of knowledge about the ADDM report and about the rising rates of autism. So thank you so much.

Walter Zahorodny

I am very grateful for the opportunity to discuss our work and to emphasize that the alarm is out there. The alarm started to go out as early as 2010, and hopefully by some luck or happy constellation of forces, we can get this to be treated seriously as the important phenomenon that it is.

Rebecca Estepp

That’s my hope too. Thank you so much.

Walter Zahorodny

I’m grateful. Thank you.

Show Buttons
Hide Buttons