INTERVIEW WITH JAAK PANKSEPP PhD.
Jaak Panksepp, Ph.D. is one of the leading theorists
and researchers in the areas of biochemistry and Autism.
Much of his work has focused on studying the brain mechanisms of
emotionality and developing animal models of Autism and other childhood
disorders. His research on beta-endorphins and naltrexone is a major
contribution to both understanding and treating individuals with
Autism. Dr. Panksepp (JP) was interviewed Dr. Stephen M. Edelson
(SE) on March 11, 1997.
Could you describe your background?
My interest in this area goes back to graduate
school when I was eager to be a clinical psychologist. As an undergrad
at the Univ. of Pittsburgh, I worked in a child development clinic
and the back wards of a mental hospital for several summers, and
I became fascinated by these types of human problems. The real eye
opener was when I went to graduate school at the University of Massachusetts.
I realized that to be a clinical psychologist, more than anything
else I would be learning a variety of pencil and paper tests that
really did not relate to the human problems that I had encountered
in my previous work. I had become interested in the severest forms
of mental disorders but in grad school it did not seem I was going
to be exposed to much more than the garden variety of everyday neuroses
I decided that the level of knowledge was so meager
in clinical psychology that I would be better off doing something
else. Since I was much more interested in understanding the biological
depths of human emotional problems, I decided to pursue psychology
from the neuroscience side. I made one of those educational shifts
that is quite rare these day - from clinical psychology to behavioral
brain research. I became especially interested in fundamental brain
functions - the basic emotions and motivations - that evolution
has constructed in our brain rather than those that emerge from
learning and individual experiences.
I have never regretted the move, and my impression
is that the discipline of clinical psychology has changed very little
in the intervening 30 years while brain research has undergone a
remarkable revolution, with lots of useful new knowledge. At this
point I strongly believe that all psychologists should be well trained
in the brain sciences. Regrettably, they still are not.
Most psychologists are still working on 'top down'
verbal-semantic approach to understanding the mind. I prefer the
bottom-up approach. Many others feel they will understand basic
mechanisms of mind by studying computers; but at some point, these
strategies becomes comparatively unproductive. I think you have
to tackle the nature of the organ of the mind rather directly, and
it is remarkably difficult to really study brain functions at the
human level in sufficient detail, even with the recent array of
spectacular brain imagining technologies. Many of these techniques
give only very global answers, and they can misinform one as easily
as inform us about the many remaining mysteries of brain functions.
This makes behavioral brain research one of the most important areas
of psychology. But it remains one of the smallest areas. In any
event, without incorporating brain research into psychological inquiry,
psychology cannot become a solid science.
B.F. Skinner has always stressed that the underlying cause of
behavior is psychological; but since our understanding of the brain
is rather limited, we can only objectively measure behavior and
the surrounding environment.
Well, certainly, but the brain also has many
intrinsic functions that we have to infer from those behavioral
studies. Skinner, who made so many seminal contributions, did little
to promote such inquiries. As you know, research on genetics and
personality has indicated that approximately half of our abilities
are due to genetic/biological sources and half to social/environmental
Although it is next to impossible to disentangle
these influences, scientific psychology of the 20th century has
largely focused its resources on studying the social/environmental
variables while neglecting many of the innate potentials of the
brain. During the tail end of this century, we have finally started
to acknowledge the evolutionary nature of the human mind, and we
are beginning to conceptualize the many intrinsic functions of the
brain provided by nature. We are beginning to realize that behavior
is only the tip of the iceberg, that it is controlled by many intrinsic
brain functions that are only refined through nurture.
We are finally beginning to realize the nurture
only refines and expands upon the many functions provided by nature.
For instance, many of the emotional potentials of the nervous system
are intrinsic to the organization of the brain, and if we do not
understand those ancient value systems, we cannot have a satisfactory
scientific understanding of behavior. For the past two decades,
my aim has been to facilitate the development of a cross-species
"Affective Neuroscience" that can tell us how internally
experienced feelings, the psychological essence of our evolutionarily
provided value systems, are constructed from neural activities.
Could you tell us more about your involvement in the areas of
emotions and social behavior, and how they relate to Autism?
After I finished graduate school I spent about
10 years conducting research on energy balance and sleep/waking
states. I was "paying my dues" so to speak, and I learned
quite a bit about how the brain really works. Then I became intensely
interested in trying to solve what seemed at the time to be an unsolvable
problem - understanding the neural nature of all of our basic emotional
processes. Most investigators, at least in my field, still regarded
emotions as a pseudo-problem - as empirically untouchable.
At the time I started, most felt confident that
no one could credibly address the underlying issues in mechanistic
ways, for instance, emotions as neurochemical processes of the brain.
However, in 1972, three reports were published indicating that an
opiate receptor had been discovered. Everyone started thinking in
functional terms as to what this newly discovered neurochemical
system is doing in the brain. In line with traditional medical practice,
the obvious ideas were that such neural systems controlled pain,
coughing/respiratory and various gastric functions.
However, we decided to focus on the possibility
that it was a prime mover in creating social feelings and regulating
social behaviors. Many of my colleagues viewed this work rather
skeptically, with a raised eyebrow, so to speak. Our guiding central
idea was that there was a remarkable family resemblance between
social bonding and narcotic addiction - from the initial attachment-dependence
phase to the eventual tolerance-withdrawal phases.
When we final began studying this possibility
empirically, it turned out to be a productive idea. It rapidly became
clear that when we gave animals very tiny doses of opiates, they
were not distressed by social isolation and they became comparatively
unsocial (even though could exhibit increases in certain social
activities such as rough-and-tumble play). When we gave them opiate
antagonists, such as naltrexone, they were more disturbed by social
isolation and they became more eager for gentle and friendly social
contact. It was not a far step to imagine that these opiate effects
on social behavior might reflect something that is happening in
childhood disorders such as Autism.
For quite a while, we struggled with the two logical
alternatives - whether such kids might have overactive opioid systems
or underactive ones. It is easy to build a compelling logic around
either view; but when we focused on the data, it was clear that
only the animals given opiates became unsocial and less pain sensitive.
Thus, it seemed more compelling to suggest that some kids with Autism
might also have too much opioid activity in their brain. This was
especially attractive since there were experimental drugs, such
as naltrexone, that could reduce such brain activities. Still, in
the back of my mind, I thought, and still do, that some of the kids,
perhaps the insecure/anxious ones, have too little opioid activity.
Some have suggested that our thinking was only
focused on the ß-endorphin system of the brain, but in fact we were
open to any of a large number of opioids being imbalanced in Autism.
At the present time, it is fairly certain that certain opioid systems
are imbalanced, but the classic ß-endorphin system does not appear
to be one of them. In Rett
Syndrome, however, high ß-endorphin is
present. Thus, right now we can be confident that some autistic
children do have elevated opioid activities in their bodies.
Do you have any guess why that might be? Did something happen
during the pregnancy or might it be genetically related?
I will not even take a position on the underlying
reason at this point. We are nowhere close to specifying the exact
nature of the imbalances. There are so many possibilities. There
could be a genetic flaw whereby there is an overproduction of opioids.
There may be too many receptors, or ones that are too sensitive.
Conversely, there could be a deficit in an antagonistic system that
shift overall brain balances. Alternatively, imbalanced opioids
early in development may have promoted an unusual organizational
pattern in the brain.
Also, we now know what a remarkable number of
different opioids actually exist in the brain and body. Some are
responsive to stress, most control pain, some create feelings of
pleasure, and others have no known functions yet. Yet others are
contained in dietary sources, such as the casomorphins from milk
protein, and Karl Reichelt has shown that some of them enter the
body, probably because of incomplete digestion and a leaky gut.
Most of the hard work disentangling these influences still lies
ahead, and some of the possibilities simply cannot be even tested
Perhaps different forms of Autism are expressed
through different opioid systems. Also, it is highly likely that
some forms of Autism have no major connection to the opioid systems
of our bodies. I think everyone is beginning to accept the likelihood
that Autism is a multi-factorial disorder. Margaret Bauman's work
suggests the initial problems are manifested during the second trimester
of pregnancy because of the abnormal patterns of brain development.
There is not going to be a single gene which causes Autism, not
a single brain chemical system, nor is it caused by a single environmental
insult. It appears to be the result of many converging biological
and stressful influences.
When you talk about stress - are you talking about physiological
stress rather than psychological stress?
It is often hard to distinguish the two. Indeed
stress has been hard to define psychologically, even though it is
quite easy physiologically. Biologically stress is anything that
activates the pituitary-adrenal system, and stress at certain times
of life can have very long term consequences. For instance, excessive
physiological stress responses during pregnancy appear to lead to
some forms of male homosexuality, and the critical factor seems
to be excessive exposure of the fetus to stress induced release
These opioids are released probably in the body's
attempt to counteract various stress responses, but if this occurs
during the middle of pregnancy, it disrupts the metabolism of testosterone
which normally induces male-typical patterns of brain organization.
Thus, baby boys from stressed mothers will tend to have a brain
organization similar to girls, while bodily they still look like
normal boys. This has been experimentally well demonstrated in other
mammals, and some of us believe the lessons apply to humans as well.
If one gives the opiate antagonist naltrexone to the mother at the
same time the stress is applied, so as to block the physiological
influences of the stress-activated opioids, the demasculinizing
effect of the stress is blocked.
If all this applies to humans, we would expect
that such boys would grow up to be normal men. Anyway, all these
stress effects seem to operate at a physiological level. There is
every reason to believe that naltrexone should increase the perceived
level of psychological stress in the mother. Thus, it seems likely
that it is the physiological rather than the psychological stress
response that initiates the cascade of events that impairs the masculinization
of the brain during infancy. The psychological stress without the
physiological components, simply should not disrupt normal development.
I would assume that the body reacts differently to various types
Certainly. For example, mild stress responses
are typically highly adaptive, while extreme forms can be pathological,
actually killing brain tissue. Also there is a distinct sympathetic
nervous system response pathway and a separate pituitary-adrenal
stress pathway, whereby cortisol is secreted from the adrenal gland
as the brain and pituitary respond to intense emotional events.
There are many physiological components to each of these distinct
responses, and in different situations, and different times of life,
the responses can be orchestrated in different ways. The underlying
brain systems can also learn, so we still have a great deal to learn
about the details of the underlying mechanisms.
Do you have any thoughts regarding a link between social problems
in Autism and the communication problems?
I think they are closely related. There has to
be a social motive for communication,
but its relation to language remains poorly understood. If one were
to select an brain area where social motivation for language originates,
one reasonable candidate would be the anterior cingulate area. When
this area is damaged, humans lose their motivation to speak. This
is also one of the highest brain area in which social emotions are
organized, and even though no one has looked closely, maybe autistic
people have impaired neural connections in those brain circuits.
Indeed, certain animals have wonderfully enriched
anterior cingulate areas compared to human. In whales and dolphins
this area is much larger than in our brains. If we just look at
their remarkable levels of social spontaneity, cooperation and group
coordination, dolphins appear to have more sophisticated social
emotional abilities than we do. Perhaps they can read each others
minds much better than we can. As you know, this 'theory in mind'
concept is presently very popular in Autism research. Many investigators
believe that autistic kids simply can't manage to fathom what other
people are thinking and feeling.
Obviously, if we cannot intuit the thoughts and
emotions of others, we cannot really understand what moves them
to behave the way they do. If we could represent the workings of
other minds, our desire to communicate with others would surely
be diminished, and the entire flow of communication would be distorted.
Endorphins seem to be directly linked to social behavior, which
in turn, is linked to communication as well as emotions. These three
areas - socialization, communication, emotion - can be thought of
as three core features of Autism. What are some of your observations
This whole idea of there being a brain emotional
system that control social motivation, which then provides a reason
for communication, connects up nicely with possible benefits that
can be seen in some children following naltrexone. This medicine
can increase the desire to communicate in some children. Presumably
those that are responders, which is no more than half the kids,
have high internal opioid levels which might be making them aloof
by reducing their social desires.
However, naltrexone does not activate linguistic
competence by itself. It seems to promote the quality of verbal
communication only in those children who already have language competence
but who seem to have a diminished motivation to communicate. However,
if the child is basically mute, as many are, naltrexone does not
increase language. It can increase nonverbal communication somewhat,
but by itself, it does not seem to contribute directly to language
Is there a 'best candidate' for naltrexone?
Our impression is that young children who have
some interactive skills and who have highly engaged and emotionally
supportive parents do the best. Naltrexone can decrease activity
and increase social desire, but the behaviors need to be noticed,
reinforced and hopefully consolidated via parental support. If we
simply focus on the child's traits, the number one feature which
comes to mind is pain insensitivity. This, of course, would be the
clearest indicator that the internal opioid systems might be too
However, there have been no studies that have
examined naltrexone in kids who are and who are not pain sensitive
in order to see if the above hunch really holds. However, Chris
Gillberg did measure opiate-like activities in the cerebrospinal
fluids (CSF) of a group of autistic children, and about half had
elevated opioid levels and they were the ones that exhibited the
most pain insensitivity. However, recently similar results have
not been found by a Japanese group that actually measured CSF ß-endorphin
Our impression is that adults generally do less
well than children, and this may be a real developmental effect
or perhaps older people simply have too many habits to break down.
Also, there may be dosage issues. Our experience in children is
that quite low doses (0.5 mg/kg every other day or even lower) are
How soon will a parent know whether his son or daughter is a responder
If nothing happens during the first couple of
hours after the first few medications, you probably have a child
that will not be a responder. We always tell parents to pay special
attention during the initial two to three hours after the first
medication. If the child does not have excess opioid activity, you
should not see anything at all. If the child has a high opioid level,
you will likely observe some kind of withdrawal or "come-down"
response; and this initial indication that the child might be a
responder is unlikely to be a pleasant or desired response. The
child may become mopey or tired. Some actually show some emotional
distress as if something is bothering them. a few hours after receiving
naltrexone, we say 'great.'
We believe these initial negative signals, which
soon disappear, tend to indicate that you have a potential responder,
but then one still need to find the optimal maintenance dose. Also,
it is important to remember that naltrexone is quite bitter and
needs to be disguised in food, otherwise the child may learn to
reject the medication because of taste alone.
Besides pain threshold, are there any other characteristics that
may predict responsiveness to naltrexone?
We think a kid who really likes hot, spicy or
salty foods may be a responder because this measure may also be
tapping into the pain issue. Other potential indicates may be -
if the child never cries or cries only rarely, and without tears.
The brain opioid system controls crying behavior in animals. If
you give opiates to animals, they do not cry at all. Perhaps another
feature is when the child is socially aloof. We suspect that the
more social and emotional autistic children are less likely to respond.
When you say 'socially aloof,' do you mean asocial, where they
can take it or leave it?
Exactly. They appear to be the classical Kanner-type
children. They are the ones who prefer to be by themselves and appear
self-satisfied. They are not the anxious types. If anything, naltrexone
may intensify anxiety symptoms, so one should use the medication
with special care, or very, very low doses, in those children.
Another characteristic may be those kids who are
aggressive but in a playfully nasty manner. In other words, naltrexone
can reduce a teasing type of aggression, such as pulling hair -
where the kids are doing things to provoke attention or attempting
to see how far they can go. In our animal models, naltrexone reduces
this type of play-fighting quite well. However, when it comes to
real anger types of aggression, naltrexone may actually increase
that in certain children.
Could this type of playful aggression release endorphins?
Yes, we have some evidence that rough-housing
play releases opioids in the brain of experimental animals. We can
even intensify playful kinds of aggression in animals by giving
them small injections of opiates. Also, Lorna Wing claims that if
a child desires rough and tumble play, but practically no other
type of social activity, it may indicated the presence of Autism.
Thus, the high desire for rough-housing in an autistic child might
be a reasonable indicator that naltrexone might provide some desirable
benefits. What is now very clear is that the brain's own opioids
clearly control rough-and-tumble play urges in animals, and I am
sure the lesson also applies to humans.
You were one of the first ones to speculate the importance of
oxytocin and Autism. What are your current thoughts on this matter?
It certainly appears that oxytocin is a player,
but precisely how it is involved remains unknown. Clearly, oxytocin
controls a lot of social processes, including loneliness, amount
of social interaction, motherly feelings and sexual ones as well.
However, just as with the opioids, you can play the logic in several
ways. Maybe the kids have too much or too little. Maybe it is not
levels of oxytocin, but brain receptor distributions and sensitivities.
At several scientific meetings, Hollander has reported seeing some
improvements after autistic adults received oxytocin sprayed into
their noses - a route of administration whereby some gets into the
On the other hand, a Japanese obstetric study
has suggested that the administration of oxytocin to mothers during
birth may actually contribute to autistic problems later in life.
We are presently conducting a large developmental study where we
administered newborn rat pups very high doses of oxytocin directly
into the brain for the first few days of life. We have yet seen
anything that give us any reason for concern about this causing
autistic-like behavior pattern. There were no changes in pain sensitivity,
overall activity, fearfulness, nor basic playfulness. There were
some tiny effect on competitiveness during play but the effects
were tiny. Thus, if you take willing to accept such data for what
oxytocin might do to human children developmentally, there appears
to be little reason for concern.
However, there are reasons to believe that abnormalities
of the sister hormone, vasopressin, may be more important, since
interesting brain and behavioral changes can be induced by manipulating
this hormone. Still, I have a gut-feeling that there is an oxytocin
component to Autism. The Hollander result makes perfect sense to
me based on the animal data, where increases in certain social behaviors
have seen following oxytocin administration by many investigators.
Still, I think the bottom line is bound to be more complex than
any of the simple ideas we presently have.
What are your thoughts on melatonin?
I think melatonin is certainly a god-send to
many families right now. There is no question that low doses of
melatonin can promote sleepiness and stabilize daily rhythms, but
it must be given only once a day and at the right time, which is
about half an hour before bed-time. We have seen many children where
melatonin has had remarkable benefits for stabilizing the sleep
patterns of kids, and if kids sleep well, they are bound to be better
during the day. It is unfortunate that there is no proper scientific
documentation of these issues yet, but the ready availability of
the medication allows parents to evaluate it for their own child.
This is not the case with either naltrexone or oxytocin.
Could you describe some of the other ongoing studies in your laboratory?
We are very interested in attention deficit/hyperactivity
disorders. Although there are some kids that have real brain problems,
I suspect the majority being diagnosed right now do not. They are
simply too rambunctious, exhibiting many childish behaviors that
cannot be tolerated in a well-run educational system. However, it
seems scandalous that so many normal kids are being placed on heavy
duty drugs that may have life-long consequences.
Such as Ritalin?
Yes, Ritalin, Dexedrine, and others. I think
these drugs have a place in childhood medicine - there should be
no question about that - but I doubt if they should be used as widely
as is presently being done, especially since such drugs can permanently
"sensitize" certain neural circuits, such as the dopamine
systems of the brain. What this "sensitization" means
is that a brain system become chronically over-responsive, and since
dopamine controls our eagerness, individuals that have been sensitized
should want more of everything than is normal.
From animal work, we know such creatures want
more food, more sex, and more drugs. It's very scary if these types
of changes are being produced in kids, but no one has yet really
evaluated the matter empirically Even worse, practically no one
is really talking about such issues. It may all come back to haunt
us one day. There is no question in my mind that the symptoms which
lead to the diagnosis of ADHD
are caused by the brain differences, but I am not convinced that
it is usually caused by a medically certifiable abnormality of the
brain. Often it may simply be the slow maturation of certain executive
systems of the brain such as those in the frontal lobes that give
us the ability to have foresight and self-control.
Anyway, we are beginning to model ADHD in animals.
The best MRI data today suggests that these kids have slightly smaller
frontal lobes on the right sides of their brains. We have recently
found that we can produce striking hyperactive symptoms in laboratory
animals by damaging one their frontal lobes, and it does not matter
whether it is on the right or the left. We still need to determine
whether Ritalin and Dexedrine can reduce hyperactive symptoms in
these animals. The bigger question in our minds is whether other
treatments might do the same. One approach we want to eventually
pursue is to see whether the excessive activity can be controlled
by various patterns of dietary ingredients and forms of bodily exercise.
Do you know about Mary Coleman's study? She investigated the effectiveness
of Ritalin as well as vitamin B6 on hyperactive children. One group
was given Ritalin; a second group was given vitamin B6, and a third
group was given a placebo. Both the vitamin B6 and Ritalin groups
improved significantly as compared to the placebo group, and there
was no difference between the Vitamin B6 and Ritalin groups. The
study was published in Biological Psychiatry, 1979.
That's marvelous. We should also try to evaluate
amino acid precursors for dopamine, such as phenylalanine and tyrosine,
as well as those that promote glutamate activity in the brain. Another
direction we hope to take is to see if rough and tumble play can
reduce ADHD symptoms. I think half of these kids diagnosed with
ADHD are simply highly energized kids who need to blow off some
playful steam early in the morning before classes. It's amazing
that no one has yet tried an intervention as straightforward as
I would not be surprised if play activates various
"growth factors" in the brain such as BDNF (Brain Derived
Neurotrophic Factor) so that lots of normal play would have long-term
benefits on the development of the brain. In this same vein, I would
not be surprised if many of the problems of Autism emerge from deficiencies
of certain neurotrophins. We can now produce animals missing certain
growth factors, and they exhibit specific deficits in hearing, vision
and touch - in many of those systems where problems are seen in
Very interesting. Do you have any final comments?
Just to come back to one of my pet peeves - that
psychology needs to become biologically-oriented. We have this wonderful
discipline, but one where people seem to be doing the same thing
over and over again, especially with the many verbal and pencil
and paper approaches. Psychology departments have very few people
that really understand the way the brain operates, and that is a
great shortcoming of the area. If we all knew much more about the
brain, we would be able to help kids with problems, as well as adults
with problems, much better. We could finally become a solid science.
Thank you for the time, and keep up the great work!
Thanks Steve, and I am sure you will do the same.
Copyright The purpose of this copyright is to
protect your right to make free copies of this paper for your friends
and colleagues, to prevent publishers from using it for commercial
advantage, and to prevent ill-meaning people from altering the meaning
of the document by changing or removing a few paragraphs.
Reproduction kindly allowed by www.autism.org
Visit their site for more useful resources.
Click here for the full range of Asperger's and autism fact sheets and personal stories