-- - Lecture series in
educational neuroscience. This Lecture Series aims to honor both world-renowned scientists and also helps form a bridge
between scientific communities here in Gallaudet in Washington
D.C. and across the nation. This series is hosted
by Gallaudet University and the PhD and educational
neurosciences program. And our theme this year
is Breaking Down Barriers. Thank you for joining us here today. Thank you to those of
you who are watching via the livestream online, hello. And thank you for coming
here from across the city or from other universities. Thank you all for joining us. Today we are delighted and honored to have Dr. Nathan Fox
accept our invitation as a distinguished lecturer. Dr. Fox is a distinguished
university professor and chair of the Department
of Human Development and quantitative methodology
at the University of Maryland. And he is director of the
Child Development Lab. Dr. Fox earned his bachelor's
degree at Williams College and his PhD in psychology
at Harvard University. Dr. Fox is internationally known for his important work
investigating the social and emotional development
of infants and children. His ground-breaking
research about the effects of social deprivation
on the developing child is truly one of a kind. As we'll hear from him
shortly this research project has yielded high-impact, sorry, high-impact insights
regarding the neuroscience, the behaviors, and the
mental health of children who experience social
deprivation as well as the ways that they can overcome this deprivation. As our inaugural lecture
for this year's series with the theme of Breaking Down Barriers, Dr. Nathan Fox will present on the effects of deprivation on the
developing brain and behavior, lessons from the Bucharest
Early Intervention Project. So please join me in
welcoming Dr. Nathan Fox here as our distinguished lecturer. It is our great pleasure and honor to have you here with us today. Thank you so much for coming
and welcome to Gallaudet. - Thank you. Can everybody hear me or see me? And out there in the studio
audience, as well I hope. First let me thank you
for inviting me to be the first speaker in your Lecture Series. It's an honor. I will admit something which
is rather embarrassing. I've been at the University
of Maryland in College Park for about 30 years and I've never been on the Gallaudet campus until
today I think that, right. I think that that is a sin on my part. But it also, it speaks to the
silos that we sometimes set up in science and the bridges
that need to be built. And I'm especially thankful
to Professor Petitto for inviting me and for
her establishing this PhD in educational neuroscience because I think it's really going to lead I can see already from the wonderful conversations
that I've had today with faculty and students, I can see the possibilities
for links between faculty at the University of Maryland
College Park and faculty here, students at the College
Park and students here. And so I'm hoping that not only is this the inaugural lecture
for this year's series but it's really the
beginning of communication and interaction between our two campuses. And between faculty at both
universities, who are interested in common problems having to do with language and neuroscience. So really thank you very
much for inviting me. I'm going to be talking to you today about a project called the Bucharest
Early Intervention Project. But really what I'm interested
in talking to you today is the effects of adversity on brain and behavioral development and I'm going to frame
my talk with regard to the notion that is very
common among neuroscientists, less so among developmental psychologists and child development
people about sensitive and critical periods in
development and the importance of understanding the
sensitive and critical periods as we think about
interventions and as we think about the impact of experience on the developing brain and behavior. So let me just first
start with a definition of sensitive periods, I'm
taking this definition from a wonderful paper
that a neuroscientist by the name of Eric Newtson, he published in the Journal of Cognitive
Neuroscience I think it was in 2002. I may be mistaken about that. But Eric defined sensitive
periods as time limited periods of time during which the
effect of experience on the brain is particularly strong. He also said the sensitive period or critical periods allowed
experience to instruct the neurocircutry to process information in an adaptive way and that
they provide information that is essential for normal development and may alter performance permanently and now what Newtson goes onto
do in this wonderful paper of his is he proposes a
number of different ways, or a number of different mechanisms in which brain architectural
changes may occur which underlie sensitive periods and I just put them up
there just so you know this is an ongoing debate in
the neuroscience literature of exactly how sensitive
periods emerge, how they close, and as I'll mention now
some really exciting work in how they may in fact reopen. So let's go back in time, the real, in terms of psychology the notion of sensitive periods was first
discussed by ethologists, Konrad Lorenz who talked about imprintment there you see him with baby ducklings and I'm sure you all
know that story about how the ducklings imprinted to Lorenz because when they first hatched, he was the first thing that they saw move. Now, really in terms of
neuroscience, the work, the Noble work of Yubel
and Vezel was really the first to really
demonstrate the presence of sensitive periods particularly
for the visual system. You may remember that they
worked with kittens and cats and showed that the effects
of early visual deprivation in striate cortex, so you can see there at the top in Panel A, you
can see there that there's a normal striate cortex
for normal binocular vision and you can see the
disruption in striate cortex that occurs when you have
a monocular deprivation they did this both in kittens
and in non-human primates to demonstrate the effects of timing of that visual experience
on the development of these brain areas that are associated with visual perception. It's interesting in
terms of human work that a developmental psychologist by the name of Daphne Maurer who is
at McMaster University, Daphne took an experiment of nature. These were infants that were
born with bilateral cataracts. And at the time that she
first started her study, the age at which these
infants received the surgery to remove their cataracts varied because neuropathologists
hadn't really perfected the surgery so she could
find a population of infants from whom that surgery
occurred at different points in time over the first year of life. And she reasoned that
that's very much like the deprivation that Yubel and Vezel had experimentally
manipulated in the kittens and she wanted to see
whether or not the affects of this deprivation had an effect upon the infant's visual perception
and in fact she found it did. What's more interesting
or what's as interesting is she found that these what she called in a wonderful paper
sleeper effects which are the fact when she followed
these infants into adolescence, what she found is that
the early experience or the early deprivation that they had experienced
established a neural substrate or the lack of a neural substrate for these capabilities that were only able to be assessed later in adolescence. And she found that in fact
14 years later children whose cataracts were
removed for example late in infancy were deficient in certain types of phase processing something
we couldn't have known had we not followed these
infants into adolescence. They call that the sleeper effect. Now, Newtson who I had
mentioned to you earlier, he worked with barn owls
and manipulates the auditory and visual experience of
these barn owls identifying the sensitive periods for the
inputs that the barn owls need as they develop and integrate this auditory and visual information. And there's an area of the
brain that he studied called the optic tectum which is the
area in the barn owls brain that integrates this visual
and auditory information. And I would love to show this slide. Because this is the way
that Newtson manipulates the visual experience,
puts these prisms on the baby barn owls to distort their visual field of information. Or he plugs up their auditory canals so they can't hear at
different ages and he finds that the timing of visual
and auditory experience is important in terms of the integration of these two sensory modalities. The other person who I just
wanted to briefly mention, I hear that she'll be
coming here to talk to you is Marina Bedny who is at Hopkins and she studied individuals
who are congenitally blind and she's found that when she
plays language to them while they are in the scanner, that in fact, they hear language in their
visual cortex not only do the language areas light up but also areas of visual cortex and
this relates to something that Yubel talked about which is the sparing, if you will, of certain brain areas
which do not receive the kinds of experiences that they expect to receive early in life. The other aspect, which
I'm sure you all know about is Janet Worker's work
in which she finds that there are timing effects
for infants being able to identify these contrasts
across different languages. And that these timing
effects disappear over the first year of life. And one of the things that Janet found was that by nine and ten
months of age, infants, before nine and ten months of age, infants can discriminate
the sounds of all languages but by the end of the first
year they are really only able to discriminate the
languages that they hear in their environment. And in a wonderful paper
that she published 10 or 11 years ago, 12
years, 11 years ago now, she proposed that there's
not one sensitive period for language. But each of the different aspects of language phonetic categories,
phonologic categories and so forth may in fact have
their own sensitive periods in which experience plays
an important role towards the development of these processes. And I think that's an important
thing to keep in mind. Okay. The final piece that I
want to tell you about is the wonderful work currently
now of Michael Hench who is at Harvard and he asks the question, can we reopen sensitive periods? And in fact, in some spectacular
work which I really won't have the time to tell you
about today, he finds that, in fact, yes, he can
pharmacologically manipulate in the mouse the areas in visual
cortex which are associated with the closing of sensitive
periods to reopen them. And that's rather remarkable sort of scientific fantasy kind of world. The general conclusion, though, that we have from the
work on sensitive periods, is at least for sensory processes, auditory and visual processes, there seems to be evidence for sensitive periods. But many of us, many of us
study more complex behaviors. And so it's not clear whether
or not cognitive or social, social/emotional behaviors
have similar sensitive periods as do the sensitive, sensory processes that I've just told you about. Most of the information that
we have about the presence of the importance of timing is inferences that we have from longitudinal and correlational types of study. So with that in mind we approach the Bucharest Early Intervention Project and let me tell you a
little bit more about that. And you'll see why it frames this question of sensitive or critical
periods in development. Okay. So first some background. The first is that what about
the history of Romania? Romania was ruled by a Communist
dictator Nicolae Ceausescu from the mid '60s until
when he was deposed in December of 1989. He believed that the way
to increase human capital in Romania was to increase the population. And so you can see on the slide there that he had a number of things that he did in order to increase human capital, which was to ban
contraception and abortion, to put a tax on families that
did not have enough children and to reward families that
did have lots of children. The results of these
policies was a wholesale, in the country a child abandonment it really became a national disaster. The families were really decimated. And people who could
not afford to take care of these many children were
told that they could put them into institutions and that the institutions would
take care of them until they were economically able
to bring the child back. So when we think about children, at least in Romania who
were living in institutions, we often called them orphans. But the definition of an orphan
is where really both parents are no longer alive. They are dead. And in many cases the infants and children that were living in the institutions in Romania had two
parents that were alive. And in some instances, those
parents had not given up the legal rights for that child. They just left them there
with the idea that the state and that the institution was
going to raise the child. And when the child was able
to, when they were able to bring the child back
into the family or afford the child back into the
family that they could come and pick the child up. Now, when Ceausescu was
overthrown in December of 1989 there were over 100,000
children that were warehoused in these institutions, as I said, poverty was the number one reason. Many people from the United
Kingdom, the United States and Canada went over to Romania
to adopt these children. But by and large, they were unprepared for the kinds of behaviors
that they would encounter. And because of international pressure, very soon after Romania
had to ban international or did ban international adoption. And so the children that were living in these institutions had nowhere to go. Now, what was institutional
living for these children like? Well there are a couple of
things to say first of all they had a roof over their
head, they had clothing, and they were fed. But it was very routinized
they went to the toilet at the same time they
were fed at the same time they were bathed at the same time. There was no training of any
of the caregivers in terms of child development or interaction. And the staff was basically negligent of the children while they were
living in these institutions. And I brought with me a video. I'll just show you a short piece of it. Which we actually took when
we first started our study. And this is in one of the toddler rooms in St. Katherine's in the institution that we were working at in Romania. Now, there are a couple of things that I would like you to note. The first is what you should be noting is the motor stereotypies
the infants are showing, this rocking back and forth. The other thing is for those of you who have ever coded social behavior, we have a code that we use which
is called aimless wandering If you go into a day care center, you see that the kids are
playing with each other. There's very little, except
aggressive social interaction that's going on among the children. That lady there is the cleaning lady, who is working with the kids. Okay. So that's the life of the
children in the institutions. And we set out as part of an effort with the MacArthur Network on Early Experience and Brain Development we set
out to examine the effects of this early adversity on
the children's behavior. And we were also interested in effecting an intervention where
we would take some of the children out and
place them into families. And see whether or not this
intervention was successful in ameliorating in some of the
deficits we expected to find. But the third piece of our initiative, which is the frame of my talk,
is we were also interested in seeing whether or not the
timing of that intervention, whether there were sensitive
periods during which time the effects of that intervention had a greater effect than at
other periods of time. Which would be evidence
of a sensitive period for a particular domain. Okay. So here is what the
design of the study was. There were six institutions in Bucharest. We assessed the children who
were under the age of two. Who were living in those. We had criteria for who
we were going to assess. We did not include children
who had special needs or Down syndrome or cerebral
palsy once we had assessed them we randomized these children to one of two conditions, one group of children were taken out of the institution and
each child was placed into a family that we had recruited and selected and interviewed. The other group of children
remained where they were. We didn't randomize to
go to an institution they were living in the
institutions at the time so they just remained where they were. We called that group of children the care as usual CAUG we called the
other group of children the FCG, the children who were placed into family care or foster care. We also recruited a
community control group. These were children who had
never been institutionalized, who were living with their
own families in Bucharest. We called that the never
institutionalized group or the NIG and what we did was followed these children prospectively over time and we had multiple domains of assessment. There's a long list of
what those domains are. I'm not going to be able to
tell you obviously about all of these domains. I'm happy to talk about
some of them during the question and answer period. But you can see here that
there's a long list of domains. In part that was because there were three principal investigators. And we all had our favorite
domains of assessment that we were interested in. And in part, quite
seriously, we wanted to be as thorough and
comprehensive as we could be. Because we really didn't
know which domains would be most affected by the intervention. And we also really didn't
know about sensitive periods during which the intervention might have the greatest effect. And so while we had
some notion about that, we wanted to cover the range of domains. And as you can see there,
it went from everything from physical development all the way down to social competence and psychopathology. So the general hypotheses
of the study were that institutional rearing would have a profound effect upon
children's behavior, their cognitive and
social/emotional development. That removing the children
from the institutions and placing them into
families would remediate some of those deficits, some, if
not all, of those deficits. And that the age or timing
of placement would be a significant factor in explaining
the intervention effects, though, as I put up
here, that may function, that may be a function
of a different domain. So for some domains there
may be one sensitive period and for other domains
there might be another. Okay. Now I'm going to go on to the data. And I'm going to present
to you some of the findings that we have accumulated
over these past years. In terms of looking at
each of these domains. The first domain that
I've chosen to look at, or two domains are language and cognition and I'm going to first start
with language, although, I will say that our assessments
of language were not as sophisticated kinds of assessments that you probably do here. We looked at language in
both the foster care group and the care as usual group. And we looked at the
age of placement into, the age of foster care placement. So this may not be very clear. It's the Reynell scales of expressive and receptive language. And the gray bars are expressive language. The red bars are receptive language. The zero line is normal. So that's the standardized,
what a typical child would be for that particular age. And what you can see
there is that if you were in the foster care and you
were taken out before 15 months of age, then your expressive
and receptive language was non-significantly different
from typical children. That was somewhat true but less so if you were taken out
after 16 months of age. By 24 months of age,
it was no longer true. If you're taken out at
24 months of age or more, you're very much like
the children who remained in the institutions in
terms of significant delay in both receptive and expressive language. Now, when we looked at
mean length of utterance at 42 months of age, we found
another sensitive period here. Here is the community controls. These are for words and morphemes. Here are foster care children who are taken out before 24 months of age. No different from the community controls. Here foster care children who were taken out after 24 months of age. And the children who were randomized to remain in the institution, they are no different from each other. But significantly different
from the other two groups. And finally in terms of
language at eight years we had an assessment of reading. And what you can see here
is again, a timing effect. These are the children who were randomized to remain in the institution. These are the children who were taken out but after 24 months of age. And in the sort of pale green, those are the children who were taken out and placed into families before 24 months of age. So there's a timing effect here with the children being
taken out before 2 years of age looking more typical like the communities controls compared to even the children who were taken
out but after two years of age. If we look at IQ, we are beginning, we're going to start
seeing a similar theme. So here are the IQ scores
of the infants while they were still living in the institution. We used the Bayley Scale
for mental development, which is mean at 100,
standard deviation of 15. But you can see there in green are the community controls have a mean of 100 which is what you would sort of expect. You might expect it to
be a little bit higher. But it's mean of 100. So they are doing fine. But all of the children who
were assessed who were living at the time in the institution their mean was 64 and a
standard deviation of 15. These kids are seriously delayed in their daily MDI scores
at this particular time. Now, remember, all of these
infants are currently living in the institution. If we now look at the intervention effect, here is what we found, we
found at 30, 42 and 54 months, if you're taken out of the institution and placed into a family,
you have a higher IQ score than if you are randomized
to remain in the institution and you remain in the institution
I will say that unless I tell you differently,
all of the analyses that I'm going to be presenting to you are with an approach that's
called intent to treat. Now you may be familiar with it for those who are not let me tell you what it is, it's something which is
used in clinical trials. Because what happens is
you randomize participants to one group or another. But over time, they may leave that group. When you do your analysis,
though, you can treat them in an intent to treat analysis as if they still remain in the same group that they were originally randomized to. And that's important. Because most of the
children who were randomized to remain in the
institution were no longer there after about four or five years. And some of the children,
who had been randomized to be placed into our
foster care families left and were not there, either. So if I show you intent to treat data, it's as if they still
existed in the same groups that they had been
originally randomized to. So this answers the first question. It says that, yes, that there is an intervention effect up
through 54 months of age. I should mention that we
supported the families that we had selected and
that we assigned each child to up through the child
being 54 months of age. We had social workers and gave them material support
and financial support. All of that stopped when the
child was 54 months of age. And then the Government took over the support of that family. The next question that I'm
going to ask of these data is whether there's a timing effect. Here is that data. And what you can see is
that there clearly is. Right? So here are the cutoffs. Here are the children taken off. This is just within the foster care group. Here are the children taken
out below 18 months of age. 18 to 24 months of age. 24 to 30. And older than 30. And what you can see here is that there's a clear break point
at around two years of age. So if you're taken out of the institution and placed into a family
before two years of age, your IQ is in the normal range. If you're taken out
after two years of age, then your IQ is significantly lower. So again evidence for a sensitive period. Now, when we follow these
kids up at age eight, here is what we found. We found that we lost the
full scale IQ of that. That in fact, there was
still an intervention effect for verbal processing. But not for any of the other subscales and not for full-scale IQ. The intervention effect had disappeared. And there was no timing effect. So that was sort of discouraging for us. What we did is we broke,
this is intent to treat. We broke intent to treat. And we looked at the children
who remained in our families, we call that MAC foster care because the MacArthur Foundation funded the initial part of our study. So those are the kids in green. There were kids that were placed
into Government foster care and kids that remained in the institution. This is at age eight. You can see there are very
few kids who were still in the institution. But you can see there that the children that remained in our
families over this period of time had higher full-scale
IQ processing speeds working memory and verbal processing. So the lesson here is that it's not only the intervention, but
it's also the stability of placement in that
intervention over that time that's leading to the enhanced IQ scores. But not to be deterred,
we just looked at our, actually it's just coming out
in developmental psychology. We looked at the 12-year
IQ scores of the sample. And what you can see there
is that remarkably ten years after randomization, and eight years after the we stopped supporting
the families, there still is, and this is intent to treat. There still is a significant effect of the intervention on IQ in our sample. And since we're interested
in intervention, we looked at sort of
the pattern of stability of these children over time. So that's starting at
30 months and going all the way through 12 years. And you can see that for
the foster care group, there were two groups. There was a group of children
that remained high over the entire period of time. And a group of children
that really remained low. If we look at the care
as usual group, those are the kids that were randomized
to remain in the institution, there were two groups, both of which sort of showed decreasing IQ scores over time. One that started a bit
higher than the other. But also interesting, there
was a group of children that actually showed increased scores over that period of time. And if we look at what
are the predictors of, within the foster care group of the children who remained
high and stable over time, it was attachment status of four at 42 months which I'll tell
you about in just a second. Age of placement. So evidence for a sensitive period there. And number of disruptions, number of disruptions means how many times was the child moved out of
the initial placement. And we know from the
foster care literature here in the United States
that the best predictor of bad outcomes for
children in foster care is the number of times that they moved from one place into another. And these data just echo
that in our Bucharest sample. So this sort of just summarizes that. And says that remarkably ten years after the intervention began there's
still positive effects on IQ. I know I'm going quickly. But I want to get through
a bunch of data for you. So now let me talk about attachment. We did the strange situation procedure on the children while they were
living in the institution. All of them. Now, I don't know if you all know what the strange situation procedure is but it was developed by Mary
Ainsworth many years ago and it involves the child
being observed during a series of separations and
reunions with their caregiver, their mother in Ainsworth's situation or their attachment
caregiver, and a stranger. So you may ask yourself,
well, for the children living in the institution, who was their mother? Who was their caregiver? And the answer is we
did that by committee. We gathered all of the
workers, the caregiver workers. And we said, okay, here
is Child number one. Who is that child's favorite caregiver? And that's, whoever was nominated, that was the child's caregiver for the strange situation procedure. And we videotaped them. And when we force classified them, we found that the majority of the children in the strange situation
had very disorganized kinds of behavior and attachments
to that caregiver. And indeed, the people
who coded these videos, we sent the videos to the
Mecca of attachment coding to the University of
Minnesota, Betty Carlson was the primary coder. And I remember she wrote
us an email and said, look, I've been coding strange situation videos from all over the world. I've never seen any behavior like this before I don't think we should
classify these kids this way. I think what we should do
is have a continuous score. And we should say if there's any pattern that's recognizable, we'll give them five. And you can see here one through four are these sort of odd behaviors
that have never been seen before in this strange situation. And when you look at the
data this way, what you see is that 100% of the community
sample is getting five. But only three percent of
the infants that are living in the institution have any discernible recognizable patterns of behavior in this strange situation. Bowlby who wrote about attachment
really didn't even talk about not being attached. He talked about variations in attachment as did Ainsworth but this is really saying that these children had not been able to form any kind of relationships
with an adult human. Okay. So I'm asking two questions The first question, was
there an intervention effect? And the next question is,
was there a timing effect? So in terms of their, the first question, is there an intervention effect. The answer is yes. You can see there that
in the institution sample this is at 42 months of age that there's a significantly higher incidence of insecure attachment
whereas in foster care it's now about 50/50 and in the community you have much higher incidence of security and attachment than not so there's an intervention effect both for security of attachment and also of
organization of attachment. The second question is,
is there a timing effect? And the answer there is yes, also. Very similar to the data that
I showed you in terms of IQ. What you can see here is
that there's a cut point. So that if you're taken out
before 24 months of age, you're more likely to
have a secure attachment. If you're taken out after
24 months of age and placed into a family, much less likely
to form a secure attachment. The other thing that we did
is we also looked at what is, was a very commonly reported behavior. It's called indiscriminate behavior. These are children that
will walk off with anyone. It's reported in the
literature that children who have experienced
institutionalization will do that. So we developed an instrument
called stranger at the door. I won't have time to go into it. But basically we wanted to see whether the child would go off with a stranger. And what we found was that if you were a child who was randomized
to the institution you were much more likely
to go off with a stranger than if you were in the family care or obviously the community controlled. And if we looked at
timing, we find that within the foster care group, those
children who were taken out before 24 months of age, that
indiscriminate behavior all but disappeared once they
were placed into the family. It took longer for children who were placed after 24 months of age. Okay. The final piece in my
social/emotional is about confidence. We were interested at age 12 in asking who is a confident child. And we used all of these domains
as you can see everything from risky behaviors
to academic performance to talk about confidence. And you can see there that a high score, which is what the community
controls had signify that you're a confident child at age 12. And what we find is an
intervention effect. That about 50% of the foster
care children are showing these confident behaviors. Versus only 21, 22% of
the care as usual group are showing confident behavior. And if we asked whether or
not timing is a factor here, the answer is yes. Here the cut point is 20 months of age. So if you're placed by age 20 months, you're much more likely to be evaluated as a confident child than if you're placed after 20 months of age. Okay. One of the things that we did in the study was also look at the brain. And I'm not going to have too much time to tell you about that. But I want to go through quickly some of the data that we have from that. We set up a laboratory with
the ability to measure EEG, electroencephalogram and
ERPs in the institution. And when we looked at the children while they were still living in the institution, what we found is that this is
the typical community sample. They are showing a pattern
of high alpha activity, which is what you would expect. But it's at if someone
had taken a dimmer switch and turned down the electrical activity of these children in this alpha range. It's almost absent to that. Now, we did not see an
intervention or timing effect until the children were eight years old but by eight years old here
are the community controlled. Here are the children taken
after 24 months of age. There are the children
from the care as usual. And there are the children taken
out after 24 months of age. And what you can see
there is there's not only an intervention effect but
there's a timing effect, as well. Remarkably, this effect
of intervention persists through age 12. Although, we lose the timing effect. So even though the foster
care children are showing the more typical pattern of
alpha activity across compared to the care as usual
children, the timing effect that I showed you now at age
eight is no longer there. We were also fortunate enough to do MRIs, structural imaging. And that allowed us to look
at gray and white matter and also at white matter
tracks in the brain. And just to show you, there
was no intervention effect for gray matter. At all. This is the institutionalized,
the foster care compared to the never institutionalized
no intervention effect and obviously no timing effect. For white matter, there was a
hint of an intervention effect with the foster care children. Sort of in between, between
the community controlled and the institutionalized sample. And those differences were in an area of the brain called the corpus
callosum which is the area that connects the white matter tracks that connect the two hemispheres. We also did something called
diffusion tensor imaging and I'm going pretty fast
but we were interested in measuring white matter tracks. And there are two measures. I'll be happy to talk
about them in the question and answer period one is factual (mumbles) and one is mean exclusivity
and what we found was in fact there were intervention effects that the foster care children
had white matter tracks that were more similar to the
community controls compared to the children who were randomized to remain in the institution. And this is intent to treat here. So these are the conclusions about brain. And you can see here that
it had a differential effect on gray and white matter. That some white matter tracks
seemed to be remediated. And the bottom line message
here is that early neglect what these children are experiencing
in the institution lead to dramatic changes in
brain structure, white but not gray matter showed
improvement with family care. And we have also looked at the links between the brain changes and behavior. There's some evidence to
say that these brain changes are associated with the incidence of ADHD, Attention Deficit
Disorder in the children. Although one of the things
that we're very interested in now, looking, is whether
or not these brain changes are associated with executive function. Okay. My last area or domain
is really on some level the most mechanistic I
think in terms of thinking about how can the early
adverse experiences affect both brain and behavior
in the way they have in the way they have in
the way I've showed you. And the area I'm going to
end with is about stress. And stress responsivity. And we know that disruptions
in stress responsivity from animal work in particular are thought to be essential mechanism
by which exposure to early life environment
influences human development. We have evidence that
caregivers play a critical role in buffering stress early in life. And that early regulation
of stress responses may have a lasting response on the stress system as it develops over childhood. So we assessed stress
reactivity in our sample. We did a number of tasks I'm really going to be talking about
the social stress test. But we measured autonomic activity across the different stress
paths as well as cortisol. We did blood pressure, heart rate, and heart rate variability. And something called
impedance cardiography which allowed us to look
at sympathetic activity. And we also measured we
also collected saliva and got cortisol, which
is a stress hormone. Elevations in cortisol are associated with more stress responsivity. So the question is, how does the early environment
shape autonomic activity and reactive activity
as I mentioned I'll only be telling you about the Trier test. But Trier test had a number
of different parts to it. Preparation. They had to give a speech they were told they had to give a speech
they would be evaluated for this speech, they
were criticized while they were giving the speech. Then they had to have this
math test in which they had to count backwards by seven as fast as they can from the number 1,073. So it was pretty stressful
for the kids and all this time we've measured
autonomic activity and we've taken saliva
to look at cortisol. Here is what we found, here
is heart rate I'll show you the autonomic measures but
they all show the same pattern. Here is speech preparation
here is speech test and math test the dark bar
are the community controls. The gray bar are the children intent to treat now who are randomized
to the foster care group. And the white bar are
the children randomized to care as usual. And what you can see there is that if you were a community
control and you were put into this Trier stress
test, you're stressed out. You show elevated heart rate during the speech and the math parts. If you are a child who
is randomized to remain in the institution, this is now age 12. So remember, the majority of the children are no longer living in the institution. But we're still counting
them as if they do in this intent to treat analysis, you are stressed hyporesponsive. You're not showing stress
response in heart rate at all. If you're in the foster care
group, you're in between. You're showing something
of a stress response, although it's not as strong
as the community controlled. That's true for heart rate. It's true for blood pressure. Systolic and diastolic blood pressure. It's true for sympathetic activity. That's this measure of
impedance cardiography. Here are the community controlled. They are showing the greatest increase in sympathetic activity
which is a measure of stress. If you look at cortisol,
the solid lines are the community controlled. They are showing a
typical cortisol response to this Trier stress test
in elevation with increase in stress and then a decrease. And you can see that the two groups, the two institutionalized groups, even the foster care kids, are showing this blunted response. But wait. If we look at timing,
if we look at the timing in which the foster care
children are taken out of the institution, what do we see? We see that the children
that are taken out before 24 months of
age, I'll go backwards, see if I can go backwards here, they look, they are showing a normal
cortisol response compared to the children taken out before. Let's see if I can go back. Here the community controls. That's their solid black line. Here are the solid black
line are those children taken out before 24 months of age in terms of their stress response ten years later. What you can see there they are showing a typical response not
only the intervention but also the timing of that
intervention is important. And that was also true. So stress, psychosocial deprivation is associated with blunted
physiological response, random assignment, high quality family care
following institutionalization. Mitigates the otherwise persistent effects of early psychosocial deprivation. And earlier age placement
leads to normalization of cortisol and vagal
engagement during social tasks. Okay so these are the
summary of the findings. So let me go to summarize
them in a table for you. I started out by talking
about sensitive periods and I'm going to end
with sensitive periods and I'm going to show
you the following data. I'm going to show you where
we found sensitive periods in these very complex domains
then I'm going to show you the domains where there
was no sensitive period. And the domains unaffected
by the intervention and I'll end with some
thoughts about why we see the pattern of results the way we do. So first why is it that
these timing effects have the impact that they do? So what is not happening in the
caregiver-child relationship to each of these domains? How much of the effects
come from this rotation of caregivers and not
having a permanent caregiver how much of the effects are general lack of stimulation and interaction. And how much are the effects
from absence of holding, lack of response to distress, lack of reciprocal
interaction, all of the things that infants typically expect in terms of their social interaction
in the first year of life. Here are the domains with
the sensitive period. Here are domains with no sensitive period but intervention effects. I told you about some of them. Right? Here are the domains that
are seemingly unaffected by early adversity we are talking about face processing earlier today. We assessed face processing
in these children. At each of the different age points. But we found no evidence
that face processing, emotion recognition, face recognition, judgment of attributes
about faces had any, was affected by early adversity. We also found domains
that were unaffected by the intervention, particularly
Attention Deficit Disorder, externalizing problems. Executive function. And gray matter. And I must tell you that the
executive function finding is very, well, troubling is not the word. But we don't really totally understand it because executive functions are thought to emerge from prefrontal cortex. Prefrontal cortex seems to be developing in early childhood through
middle childhood some people even say up through adolescence. And that's when you measure
executive functions. You see that curve. Yet children who had this
early adverse experiences seem to be significantly affected in terms of their ability to do executive functions. So here is what we don't know. Truth in lending, truth in transparency. We don't know why some domains like IQ and EEGs showed timing effects early on and then those timing effects disappear. We don't know I just
mentioned why some domains like ADHD or executive
function are unaffected by the intervention. And we don't know why some domains like face processing appear to be spared but what we do know is that
early adverse experiences like neglect have a profound
effect upon multiple domains of behavior and that important
domains like attachment and social relationships
are significantly remediated if children are taken out early and provided with family
kinds of environments. So overall conclusions, children raised in institutions demonstrate
significant impairment, which I told you that. And we demonstrate
specific cognitive deficits that are associated
with those experiences. Insofar as we've been able
to look at our data our model of foster care as an intervention appears to be effective in ameliorating many of the negative consequences and some aspects of brain structure and function, particularly white matter and EEG, are remediated in children placed into foster care whereas others, particularly gray matter are not. Okay. This is, I did not do all of this work. By myself, I have two
wonderful collaborators, Charles Zeenah, who is
a child psychologist at Tulane University and Charles Nelson, who is at Harvard Medical School and Children's Hospital Boston. We were supported in the initial part of the study by the MacArthur Foundation and for the last seven years
we've had funding from the NIH, National Institute of Health. Now you might ask yourself
why is the National Institute of Health interested in the affects of institutionalization in Romania and really the answer I hope
that I have convinced you is that this is really
more about the effects of neglect on the developing
brain and behavior. Neglect is the most common
form of maltreatment in the United States
to the extent our data can inform policy not only worldwide with institutionalization
but with neglect here in the United States, I think that the research will have some impacts. So with that I will thank
you for your attention and hope to get some questions. (applause) - Thank you, that was amazing. And the number of
different areas of research that you've combined into this one project is just astounding. I want to welcome everyone to
come up and ask a question, give a comment. So please. Let us know what you think. Yes, please come up here. - Thank you for your talk. It was very interesting. It made me think of another recent study from a colleague who
studied the effect of SCS on math and arithmetic. We know that SCS has a lot of, has the strongest impact on math outcomes in children from low
SCS and what they found is that they studied children adolescents, in simple arithmetic tasks and compared low SCS with children from better homes they found
that performance was the same. However, the brain networks recruited by the two groups were
substantially different. So higher SCS kids were
recruiting verbal strategies, which are thought to be more
efficient for arithmetic, at least the ones that they were testing. Whereas low SCS were using more facial and quantity processing so at that level it appeared that both
groups behaviorally were the same but from a neural
network, brain network they were substantially different and maybe I was thinking
whether something along these lines could have affected
the non-difference observed in IQ in your 12-year-olds testing, the two groups were the
same at 12 years of age. - No actually we had an intervention effect at age 12. - [Questioner] But it was, wasn't it that there was no timed intervention. - The timing, right. - [Questioner] I was wondering if that -- - Yeah, that's a very good point. So yes, so that's a great point about why the timing differences may disappear. Because really what you're
saying is something we know from the neuroscience literature is there are many avenues to
get to adequate performance. Some of those avenues
involve compensation sayings or development of the
compensatory circuits. So it is possible that
some of the children or the children who were
taken out later who earlier had very poor performance, that they have developed
these compensatory strategies. And compensatory neural
network which allow them to perform adequately. It's just that the way
for them to get there is different from the children
who were taken out earlier and more able to develop
these adaptive strategies. Now, what that speaks to is the need to have finer measurements. Not only in the brain but
also finer measurements of behavior so that we can break down, so IQ is sort of a general
large undiffused concept and really what we need is
much finer behavioral analysis as well as neural analysis
that would allow us to break down what are the
individual competencies or the individual brain
networks that are associated, much like the study you
have just talked about, what are the individual networks or the individual behavioral competencies that allow for performance. And I think there you're probably right we would still see differences within the foster care group between
those taken out before and after 24 months. It's a great point and we need to be able to do that in order to
answer our own questions. Thank you. - Hi there. I'm the adoptive mother of
two children from China. One was brought to an
orphanage at one month of age and I adopted that child at seven. The other one was brought
to the orphanage when they were about four years old. They initially lived with their family. And I can see a huge difference between my two children
not only in their behavior and their language or how
they figure things out, their reasoning. And what you have presented
helps me understand a lot about what's going on in their minds and how their minds are functioning. My older daughter, who is
currently a student here at Gallaudet, you can see
from what you've said, a lot of what you said parallels
and it helps to make sense because she didn't have
proper language stimulation when she was very young. She didn't have that sort of exposure. She had her change of
caregivers was frequent she had multiple nannies she
didn't have stability in who her caregivers
were when she was younger and you see deficits in
her ability to empathize. She is catching up. My question is will she ever get caught up with those of us born and raised in our families will she ever
perform somewhat normally? You said that if kids are
taken out roughly before the age of two that they will be fine. But sometimes you don't see
people get adopted until they are 12, 13 years old. And can they ever get caught up? And I understand that
measures like IQ are kind of ambiguous how do you measure IQ when they don't understand what's
happened prior to that? - Thank you for your comment and it's a wonderful story. The answer is that if
developmental psychologists were able to predict
at the individual level we would all be extraordinarily wealthy. But we can't. Because there are so many factors that go into the trajectories of their, to individual children and their response to adversity so our data
are at the group level. We obviously see that at the group level there are significant effects of this early deprivation and neglect. But even within the children,
I don't know if you noticed, even within the children
who were randomized and remained in the institution,
there were some children that had increasing IQ scores over time. So there are individual differences that, which we can't really identify what those individual differences are. That may in fact underlie success. And competence and adaptability of individual children over time. I don't know if that's
a satisfactory answer. - Well, as a mother,
how can I be satisfied? - Right, exactly. (chuckles) - I mean, really, yeah so essentially the point is that children can be adaptive but what I've seen is that really those who are adopted don't seem to have the greatest ability to
be adaptive just frankly. It seems like there's
a great deal of tension that's always underlying what they do and I think that's part because nobody has really prepared them, nobody has said, hey, this is going to happen
in life and when that happens, you can do this or even
if they said you might go to a new mom and dad and then if they come to the U.S. or wherever
their adopted families are I don't know that they
have been prepared for that and I wonder if that has an impact on the ultimate outcome and
if what happens early in life really plays a big role. - Yeah, I think that you're correct. Even within the United States in terms of national adoption there
are issues for children and families in terms of the relationship that they establish. Even if they are adopted
soon after birth in terms of a psychological relationship between an adopted parent and adopted child and that's all the more
complicated and complex when that child comes from a different culture and when that child has
experienced different forms of adversity. Early in life. The best thing we can say is there are important individual differences that sustain some children to
be successful and adaptive. Over time. - How do you measure that gray and white matter in their
brains when that's happening? - Well, measuring the gray and white matter is the easy part. We do something called structural MRI. We use an MRI machine. And we look at the amount of gray and white matter in their brains. But I just want to say
something that even with these differences in gray and white matter that I showed, that's not, that in itself will not
explain who is going to go on and adapt and be successful
and who will not. There are multiple factors
at the individual level that need to be accounted
for not the least of which is the kind of support the
child gets post adoption and the kind of family that the child goes into post adoption. - Thank you. - Thank you. Yeah. - That was an absolutely fascinating talk and very thought provoking
and this is more of a comment but I thought I would share
with you that you were, one of the profound relevance of this is we can gain insight
into the impact of neglect. Neglect in child development. One of the things we see,
so this concept of neglect is a very interesting
concept because that's a situation these children
were absolutely clearly in this social, socially deprived, socially attenuated neglectful context. But we see children showing
dramatic impacts similar to yours with regards to language and reading and higher
cognitive processing. When neglect is not
your classic construct. So they are from nice middle class homes. These are deaf children, typically
they have hearing parents and they are going to
get cochlear implants and these parents have the best intentions and give them all of
the Baby Einstein toys but the medical institution
has told these parents, do not expose these
children to sign language. So it's not neglect in your classic case, they are not socially neglected but they are linguistically deprived. So this is like a select
column of human interaction has been significantly
withheld so now these children are going to get a cochlear
implant at 12 months and 18 months old and if it's at 12 months it's not tuned and ready to
go until around 18 months so from birth to 18 months
there's a kind of neglect they haven't had the input. So while they had all of the
social love and richness, the human brain can be
impacted if it's just something like withholding of the
timing of the experience and exposure to language. So these children have
telltale indices of having had this late exposure to
the patterns of language that really impact reading and then cascading impacts other
cognitive growth factors. So I just find it fascinating
that neglect can be social. Or you can have social
rich, rich environments. But you don't get language and still see a devastating ripple effect. - So two comments on your comment. The first is one of the
things that I didn't mention in my introductory remarks
there's a wonderful paper 1983 in child development by a neuroscientist by the name of Bill Greenough and the paper was about experience expectant versus experience dependent learning. Experience expectant
learning or exposure meant that the brain is expecting
a certain type of input. When it doesn't get
that input, what happens there may be a sensitive period in which that input needs to occur. And if it doesn't occur
or if it occurs later, it will have cascading effects downstream. The data I showed you
from Daphne Maurer's study with children who had
different ages in which they had bilateral cataracts and then the cataracts were
removed is an example of that. And it's similar it seems to me to the data you told us about in terms of deaf children not being
exposed to language up through the first 18 months of life. Similar in the sense that
here is another domain like vision that the brain is
expecting a certain type of, a certain level of experience. And it's not getting it. Whether it gets it or not at a later age, it may compensate, but it's
going to be as in Daphne's the sleeper effects there are going to be cascading effects downstream
in terms of development. - We have time for one
more question or two, if they are short. And before we finish up
the question and answers, I want to remind everyone that we have a small reception after this. Just over in the SLCC Atrium. So just across the way. Come and join us. And we can continue
the conversation there. With some food and drinks. So one more question here? - You think you can both have questions? - We can do two. But Zac, Alaria already had a question. So come on up. - Earlier today we were
talking about some research from your project that you continue to do where you're following up with these children who I
believe you said are 16 now if you don't mind could you maybe talk a little bit about what your expectations are for the study ongoing? - Sure. So 16, 15 and 16, which is the age that we're seeing now is
deepest and darkest adolescence. And you know there are
some people that argue that adolescence is
another sensitive period in brain development. In the sense that there are experiences that are expected to occur
during that period of time. Hormonal changes that
affect brain structure and organization. So we are asking whether
or not the effects of early adversity have
cascading influences and heightened what are the
typical kinds of behaviors that are seen in adolescents. So for example, risk taking. Externalizing kinds of behavior. And other forms of psychopathology
that are generally, emerge in the adolescent period are those forms of
psychopathology and behavior, are they enhanced in
this group of children? Or are they diminished as a function of these earlier adverse experiences? And so that's one primary
issue that we are, that we're quite interested in. Yes, you had a question? Come on down. - Thank you for your talk. I might be completely off
base with my thinking here. And if so, just tell me. I'm thinking about what if
the whole thing is a defense? Meaning like say that
the brain's responses and activity in the brain is
such that in and of itself, it is a defense. So I'm thinking about if I had grown up having
primarily visual experiences. And then say that I go to a
place where everyone speaks and I can't participate in their language, I would be so completely at a loss, I wouldn't know how to interact. But if I were in the same sort of state, always having a visual experience, then it wouldn't be anything different. So do you think it's something about a deficiency or a defense? - Yeah, I think that's a great question. And I do think that there are behaviors that children who have experienced neglect and adversity and particularly
institutionalization, that they exhibit which
are, I don't even know that I would call them a defense, I would call them an
adaptation that they have to the context in which they are living. The best example, which I really
skipped over really quickly is this indiscriminate behavior. So parents who have adopted children from institutions report anecdotally that children will walk
off with anybody who gives them their hand. Somebody sticks out their hand and the child will walk off with them. They also report that
the children are not, they don't seem warm and don't bond easily with their caregivers. Now if you're a child growing up in an institution and somebody, a stranger they may be sticks out
their hand and says, come with me, if you don't go with them, you're going to be in trouble. So your adaptation to that environment is to walk off with whoever
it is who is sticking their hand out and saying, come with me. We have multiple instances
of children who have adapted their brain and behavioral responses to the context in which they are living. I don't know I would
call, you might think of it as a defensive response but really what it is is a defensive response against the adverse context in
which they are growing up. I don't think you're off base. I think it very much
sort of tells the story of why the children respond
to the context in which, in the ways that they do. - Thank you. - DR. NATHAN FOX: Sure. - This fits well with
what you're talking about. So I wanted to ask you a
question about the study on the stress where you ask the
kids through representation, to perform math and you
measured their stress response and here is the question that I had is, the kids that were in a institute had a very low stress
response is it habituation or adaptation because they
have been so much stressed in their previous past that
maybe they are not reacting it's not really a stressful
situation to them. Stranger people come by all the time or is it really a
physiological change that makes them insensitive to situations
that should stress them and therefore might
explain future misbehavior in adolescents or young adulthood. - That's a good question. We don't know the answer. But what we do know is
that that hyporesponsive, the hyporesponsivity
could be the HPA access in the autonomic nervous system involved in stress is a regulatory
homeostatic system which shuts off when it gets too high and then comes back on when it's needed. And so the dampening down and
the hyporesponsiveness could be that it is chronically been stressed. And so it has shut down. There was early, there
was work some 30 years ago on depression. That thought that individuals
who were chronically depressed that their stress system
had also shut down. There was in fact a test called the dexamethasone
repression test which was a challenge to the stress
system to see whether or not it would elevate in
response to that challenge and in depressed individuals,
it apparently did not. So it may in fact be it has shut down as a function of that habituation. One of the things that we didn't do, and that we sort of hit our
heads saying we should have done is measure stress
reactivity at earlier points in time because in doing
that we could have looked at the developmental
trajectory and the factors that were associated with
either that habituation or the shutdown. So we'll never be able
to answer that question. But it's really, you
know, right on in terms of the mark in terms of understanding sort of the mechanisms by which these children's behavior emerges. - Anyone else? Thank you for all of your great questions. We really appreciate them. Please now come join us and we can have some more informal discussions. - Thank you. - And make sure to fill
out your evaluation forms. And hand them in to us. - Thank you.
