- NF-1 AND LEARNING DISABILITIES
Martha Bridge Denckla, M.D.
-
Introduction by Brenda Duffy, Mid-Atlantic Chapter President. Martha B. Denckla, M.D.
started working with Neurofibromatosis and learning disabilities when she was at the National
lnstitutes of Health (NIH) with the National Institute of Neurological Disorders and Stroke back
in the early '80s. She is now the Director of the Developmental Cognitive Neurology Unit at the
Kennedy Krieger Institute in Baltimore. She is beginning her second five-year grant to study
NF-1 and learning disabilities (LD). Many of our Chapter members have participated in the
study, so we are happy to provide an update on the funding. Dr. DenckIa has also been named
the 1996 Neurofibromatosis, Inc. Scholar. This year she is going to be using her study in NF as
a model in several presentations to major professional audiences.
The study originally was funded by the National Institute of Child Health and Human
Development (NICHD) because they were interested in learning disabilities. Neurofibromatosis-1
is one of the three neurogenetic syndromes that we are studying. The others are Turner
Syndrome and Fragile X in girls. We are studying girls with Fragile X because in boys it is a
much more severe condition than you would call a learning disability (LD).
At the beginning of the study, instead of going out in the public domain and deciding who was
dyslexic and who was math disabled (and whatever), we wanted to look at things from the other
way around. We started from the gene and then went to the brain and then worked our way
outward to find out what L.D.'s exist in people who have a gene associated with specific findings
in the brain.
We are dedicated to finding out things that are important for NF-1. It turns out that, as is so
often the case, what we find out about NF- 1 is very helpful and very applicable to Learning
disabilities in general.
Since March, 1996 we have had three articles published in professional journals. The article
that was published in the Mental Retardation and Developmental Disabilities Review is on
reading disabilities. You might wonder, "Why do they want to have an entire article on NF- 1?"
The reason is that it maps out a very important pathway whereby people have reading disabilities.
So there are both specific and general things that are interesting about the work we are doing.
The history of this particular field is not that old. It started in '88, when there happened to be
three publications all of which basically said that there were learning disabilities in NF-1, but the
disabilities were mainly what were called non-verbal. People said they did not have reading
problems as much, but had so-called visual-perceptual problems.
The problem was they really were not sticking to the vocabulary of LD, because they did not
say that children with NF- 1 had math disabilities. They said that those with NF- I did badly on
specific tests. The one that you are going to hear about all the time is called "Judgment of Line
Orientation," (JLO) which is supposed to be a test of spatial ability. It is a type of stretch from
spatial ability to math because, although people in the field think that spatial ability is linked to
math, in fact, nobody has really proven this the way we have linked another whole set of processes to reading.
So what I want to make clear is that people called this a non-verbal learning disability, but they
were really talking about a specific type of underlying disability. They really did not link it to a
school subject. So they were not using the term "learning disability" conventionally, that is, with
respect to a school subject.
All these studies had limitations as to what the control group was and what they were
comparing the NF-1 population to. You will see Drs. Samuelson and Riccardi stated in a more
clinical way that they thought the people with NF-1 did less well in school than other members of
their families did. They were talking even about adolescents who were less academically
advanced than siblings. The doctors brought up the question of whether their patients' IQs were
also somewhat lower. But they said an intriguing thing: They felt the adults with NF-1 were not
different from other people's intelligence. Dr. Riccardi has said that he thinks that the so-called
"intellectual gap " is there in childhood but it is not there in the adult patients. My colleagues and
I have an interest in that.
You have to either praise or blame the design of that first pilot study or all my subsequent
work on Dr. Roswell Eldridge, because he is the one who pounded into my head that you had to
do a sibling case control study. We found in that preliminary study, in fact when I was at NIH
working with him, a left shift, which means a downward shift, such that the IQ was lower in the
people with NF-1 compared with their own brothers and sisters. This is Dr. Eldridge's design that
I have carried on ever since. Every comparison you are going to hear about is always some
matched pairs -- somebody who has NF-1 and that person's sibling who does not. Now if we
could get identical twins, one of whom has NF- 1 and one who does not, then we would be in
seventh heaven -that would be perfect.
Siblings are the best that you can do if you do not have identical twins, one of whom is
affected and one who is unaffected. We also found this pairwise deficit on the test called the
"JLO". We find it again and again and everybody finds it. There are really subtle neurological
signs that you would observe. For example, how long can you stand on your right foot? It is a
really quantified neurological exam. So there are subtle neurological signs, less perfect motor
control in the NF-1 person than in the sibling. It does not seem to make any difference whether
you are familial (inherited from one of your parents) or sporadic NF- 1 in terms of pairwise
deficits.
The idea of the JLO is that there are two lines. And then you are supposed to say, "Where are
those two lines?" (And please don't ask me to do it because I have a deficit on this test myself!)
The Judgment of Line Orientation graph is just to show you the visual image of matched pairs,
comparing a person with NF to a person without NF. In one family there was no difference. So
it is not a hundred percent of everybody that is worse than his or her brother or sister. Once again
the familial/sporadic distinction does not make any difference.
There is another finding in our study that is so interesting. Remember everybody was talking
about non-verbal learning disabilities. Well, we really do not find that, although we do find the
left shift.
Both parts of the IQ are different. The verbal one, however, is going to turn out to be the
most different. It is like the statement in George Orwell's "Animal Farm": "All animals are equal.
But some are more equal than others." So it really is more the verbal IQ that distinguishes NF-1.
By contrast, on the other scale, only one test is different, and it accounts for the whole nine
yards of why the performance IQ is lower. Block design is putting together a design in
conformity to a model; you have to use spatial analysis and synthesis. That is the whole story.
Nothing else here is significant on the whole performance IQ. This begins to shake you up when
you wish to call this a non-verbal learning disability, because everything else that you had to do on
this Performance part of the IQ resulted in no significant difference. It is just a whopping
difference on the single sub-test, Block Design, the most spatially loaded sub-test of the IQ test.
(Somebody could accuse me of doing research out of empathy because this is my worst
characteristic. And I do have empathy because that is also my worst sub-test.)
Note that with NF-1 there are far more verbal things that are problems -- Information --
Similarities -- Vocabulary. Not Arithmetic! In the literature on non-verbal learning disabilities,
this should be the "gold standard" of non-verbal learning disabilities, because Arithmetic was
originally what Dr. Rourke held up as being the sign post of non-verbal learning disability.
This is why the family design is so crucial. If I told you I had a bunch of children whose
Information and Vocabulary, for example, were different, you would say, "Well, information and
vocabulary, those are things kids get from around the dinner table. Those are things that come
from the socioeconomic background. "
So when you get a different bunch of controls that are from who knows where, and you say
that something like Vocabulary and Information are different, people can just laugh at you and
say, 'Well, I am not interested in that. It could just be family environment." But that is why you
use siblings. It is a much more robust and interesting finding. Kids from the same home should,
on average, not show a big difference from each other in things like vocabulary and information.
On the Woodcock-Johnson educational battery you get the same type of difference, though
not a significant difference. The Woodcock-Johnson battery of tests is used by most school
systems to decide whether you are learning disabled. The big differences are in the subtest Word
Recognition, which means "sight words." You just read from a list of words at your grade level.
That is the biggest difference. The second biggest difference is in Passage Comprehension. To
some extent, Applied Problems or word problems are an NF-1 issue. So again, math calculation
is the least different thing about children with NF-1 academically. The most different thing is their
reading, a completely unexpected finding from what we thought when we went into the study.
Another test we use is the Wisconsin Card Sort Test. The only part of this test that is different
in NF-1 is the verbal. It arrives at different categories like color, number and form. It happens
that the categories achieved score correlates very highly with verbal IQ, which we already talked
about being an issue with NF-1.
A test of attention is given (related to correlates of ADHD); no significant difference is
indicated. So there is no excess ADHD cognitive deficit in the group with Neurofiromatosis at
all.
We have a picture-naming test, Boston Naming, which is a type of vocabulary test that is part
of our standardized language measures that is an NF-1 issue. We have something that is the heart
and soul of all the research on reading disability, "phoneme segmentation". It goes like this: Say
"great." Now say it without the "t". What is left? The word is "gray" in case any of you are
struggling. Say "cup." Now say it without the "c." What's left: "up." This phoneme
segmentation test has become enshrined and ensconced as a way of testing whether a person has
the underlying deficit that leads to a reading disability. The NF-1 children have trouble with it.
As a group, the children with NF-1 had trouble with naming phoneme segmentation as well as
phonological memory, which is an even more wicked test in which you are trying to remember
nonsense. If I say to you "mug, hex, web, yag." Okay, now say it back to me. Now we are
going to do it five times and see whether you remember "mug, hex, web, yag."
If you want to reduce a bunch of reading disabled kids to tears, give them this test. Because
when you take meaning out of words, they have a lot of trouble. Basically, leading disability is
based upon people not having a good ear for the speech sounds in the language; it is the mirror
image of talent for music, as when a person has a good ear for music, or a bad ear for music. The
literature on dyslexia in general has come up with phonological "ear for language" issues, which
are the things that the children with Neurofiromatosis that we have seen have the most problems
with.
We also use a comprehension test, which requires understanding complicated commands. A lot
of NF- I language findings are seen phoneme segmentation, phonological memory and the Token
test, which taps receptive syntax. This is an example: "Before you pick up the yellow square,
touch the blue circle" you have people try to understand and carry out those commands.
These are language disorder findings associated with NF-1 and not a non-verbal learning
disability. This is the punch line of our study. There is an NF-1 related problem in only a couple
of things. Contextual vocabulary is a problem, but them is not a generalized intellectual reason
for having trouble. It is much more like a specific learning disability which is language based.
In review, the things where the NF-1 sibling has trouble compared to his or her own brother or
sister are: three of the verbal things from the Verbal IQ test, the Block Design part of the
Performance, the line orientation (JLO), which is the most significant one; the Boston naming, and Token Test.
Less prominent but still issues with NF-1 are phonological challenges and
Categories from the Wisconsin Card Sort test. Whatever Learning Disabilities risk NF-1 has, it is
more verbal than it is nonverbal. So the early conclusions about non-verbal which was largely
based on that one test, the judgment of line orientation, that is true; but you have to say you have
a set of mostly verbal deficits and one spatial deficit. But you have an awful lot of good stuff in
between, most of which is nonverbal. There are different kinds of non-verbal. In other words,
non-verbal is not one test. So that one nonverbal test is not normal but a lot of other nonverbal
stuff is excellent. It tells researchers to be more analytic about the nonverbal description.
Now, I will go on to discuss the significance of the "UBOs" (unidentified bright objects) which
are the bright signals on the MRI.
On this slide are the researchers claiming there is no UBO effect, I could even add some more
of them. Now, our paper and the one of Dr. North from Australia in 1994, are the only ones that
agree with each other. A more limited agreement comes just now from Dr., Bart/Moore. We all
believe that the UBOs, at least those in some locations, do have something to do with the
cognitive problems.
The issues about NF-1: Do the UBOs have anything to do with the IQ? Do they have any
impact on the neuro-psych test, such as the JLO, the most prominent deficit that has been found?
And, if they are significant, what happens when they disappear? One thing I can tell you very
firmly from our very first publication, when we searched the entire Johns Hopkins Hospital NF-1
patient MRI file, adults with NF-1 do not have UBOs in their brains any more. Children do.
Adults do not. Not all children do. Two children out of our group with NF1 have not had any.
Now what we tried to do was to predict, and this is tricky, (and a lot of neurologists do not
understand this). We are not predicting low IQ but rather lowering of IQ, that your IQ is lower
than your family was otherwise going to have a child exhibit. So it is "lowering". It is not being
retarded. It is asking if the Genome has all kinds of instructions for IQ, and I throw some of
these things into your brain, whatever they are, does that impact on your IQ? And in what way?
The things we also considered were: Your age. Because again, I told you Dr., Riccardi says
people get smarter with NF-1 as they get older. So we thought within childhood and adolescence
maybe they get smarter just on the basis of age and there would be less of a gap. What about the
factor, familial versus sporadic? And we'll put in as possibly important total volume, because we
were able to measure the volume of these things. The total volume of your brain that these white
plaques occupy. And then there are the number of locations of UBOs. Then we tried to say,
"Which one of these things would say how much lower your IQ would be than your sibling's IQ?
The sole significant predictor of the lowering of IQ in NF- 1 affected persons is the number of
UBO-occupied locations That is, 42% of the variance in the discrepancy of the full-scale IQ
from that of the unaffected sibling, which is the only significant amount of the variation that is
accounted for, is derived from the number of locations that have UBOs in them.
In the two cases we have in our study who do not have UBOs the NF-1 happens to be familial.
Whether you have UBOs or not, does not seem to have much to do with whether you are familial
or sporadic NF-1. In those two cases, the IQs are remarkably on the same level as their siblings.
Again "unidentified bright objects" are actually a signal rather than an object which is the original
name. (Because Dr. Riccardi is so philosophical, he does not like the term "object," so he says
"signal" because the MRI gives you a signal on a certain physical parameter.) They just look like
some fuzzy whiteness in the brain. We are now up to about 24 cases, four more than we had
available to do the analysis on, but we still have only two cases where the NF-1 affected person
does not have any UBOs. Those youngsters have exactly the same IQ as the brother or sister. So
that is kind of intriguing. We told you the parents, that the UBOs go away, so we are very
interested in the timing of when they go away and if the adolescents then get smarter. Maybe that
is why Riccardi observed adults get smarter. So, it is a real interesting story.
page 1 | 2
|