UPMC Physician Resources
Neurodevelopment of Cognitive Control through Adolescence: Imaging Studies in Normal Development
Dr. Luna uses various studies of cognition and brain maturation to help improve early identification of psychiatric disorders and improve treatment of such disorders, specifically for adolescent patients.
Upon completion of this activity, participants should be able to:
- Improve diagnostic skills for identifying early indices of psychiatric disorders including schizophrenia
- Increase patient outcomes when prescribing medication treatment
- Improve patient outcome by taking into consideration sex effects
- Geier CF, Padmanabhan A, Luna B. Immaturities in incentive processing and executive function in adolescence [Review]. Neurotoxicity Res. In Press.
- Ordaz S, Luna B. Sex differences in physiological reactivity to acute psychosocial stress in adolescence. Psychoneuroendocrinology. 2012 Jan 24.[Epub ahead of print]. PubMed PMID: 22281210
- Luna B, Padmanabhan A, O'Hearn K. What has fMRI told us about the development of cognitive control through adolescence? [Review]. Brain Cogn. 2010 Feb;72(1):101-13. PubMed PMID: 19765880.
Dr. Luna has no relationships with proprietary entities producing healthcare goods or services.
The University of Pittsburgh School of Medicine is accredited by the Accreditation Council for Continuing Medical Education (ACCME) to provide continuing medical education for physicians.
The University of Pittsburgh School of Medicine designates this enduring material for a maximum of 0.75 AMA PRA Category 1 CreditsTM. Each physician should only claim credit commensurate with the extent of their participation in the activity. Other health care professionals are awarded (0.075) continuing education units (CEU) which are equivalent to 0.75 contact hours.
For your credit transcript, please access our website 4 weeks post-completion at http://ccehs.upmc.edu and follow the link to the Credit Transcript page. If you do not provide the last 5 digits of your SSN on the next page you will not be able to access a CME credit transcript. Providing your SSN is voluntary.
Release Date: 10/4/2012 | Last Modified On: 10/15/2013 | Expires: 10/15/2014
So what I’m going to be presenting to you today is a zooming out focus from what you just heard now, I’m actually a development cognitive neuroscientist not a clinician, but I’m particularly interested in this transitional period between adolescence and adulthood in trying to understand what is occurring at the level of the whole brain using neuroimaging studies. So of course as Dr. Lewis very eloquently described to us there is a very good reason to look at adolescent, especially when interested in schizophrenia. Adolescence precedes the typical age of onset of schizophrenia and understanding the brain bases of normative development can give us a little bit of a view as to what are the vulnerabilities of the systems that might be breaking down with schizophrenia.
Now what I’m going to show you here is a little bit of more of the whole brain story that we know of changes that are occurring during the adolescent period. Now what I’m showing you here is – there we go – this is work from a colleague and friend of mine, Jay Geed at the NIH, and what he did in this study was look at the thickness of the cortical gray matter in the brain, right, this is where the neurons reside. There is a normative process of thinning of this cortical gray matter and this is a good thing. And this continues to occur through adolescence and what you can see, and I’m highlighting the adolescent period here, is that there are many regions including prefrontal cortex that have not reached adult levels. So it is important to understand that it’s not just prefrontal cortex, you can see that there are areas in temporal cortex that have to do with language, even in occipital areas that have to do with visual processing. It’s mostly the areas of the brain that support the greatest amount of interconnectivity. Also within the middle of the brain there are protracted developmental changes in the brain structure of the basal ganglia. These are areas of the brain that have to do with the motivational and arousal system. These parts of the brain in conjunction with the more cognitive executive parts of the brain are still maturing during this period.
Now what I’m showing you here is a depiction of some of the findings that have been found in the animal literature showing very provocative evidence of increases in dopamine availability that are specific to the adolescent period. This is important because dopamine is a neurotransmitter that underlies motivation but also underlies learning abilities. And this seems to be peaking in adolescence and developmental cognitive neuroscientists as myself really believe that this is actually what is underlying that increase in sensation seeking that we all are very much aware of occurs in adolescence, but might be very telling of what is occurring to vulnerabilities in psychiatric disorders including schizophrenia.
Now two processes that we know are occurring during this time specifically from histological studies is the loss of synapses and you’ve heard about this as well, David Lewis just talked a little bit about this, and the process of myelination. But when we think about this I want you to think a little bit more closely as to what would be the implications of these two processes. When we are talking about synaptic pruning we are talking about the ability to enhance the local computations the regional areas such as prefrontal cortex can sustain. We are not talking about the whole brain, we are talking about individual regions.
Now the histological studies have shown that the middle frontal gyrus which is an area of prefrontal cortex where DLPFC is located reaches adult levels by 16 years of age. The other specific regions that have been found all reach adult levels at earlier ages. Now when we think of myelination we are talking about a little bit of a different system. The myelination of the occipital tracts which continues and proceeds in a more linear fashion is actually providing support for dramatically speeding neuronal transmission. So that now you have a brain that has the capability to be behaving in a collaborative fashion and most importantly to be able to engage top down modulation of behavior. And I’ll be very much focusing on this top down modulation of behavior, which is something that we know is the first to take a toll in psychiatric disorders especially in schizophrenia.
So there are many ways to look at this. I’ve chosen to use a neuroscience approach and to look at the oculomotor system, the eye movement system to understand brain and behavior relationships for many unique reasons. Number one, this is one of the major systems that have been very well delineated at the single cell level in monkey studies so that we know its anatomy, its physiology and its neurochemistry. The tasks are extremely simple, the instructions are look at a light, don’t look at a light. So we know that there is not going to be a confound of instruction which can be very pertinent with developmental studies. The stimulus and response live in the same domain, so you don’t have to have any cross-modal translation occurring which could also engage neuro-processes that may not be mature. And also the – no spinal cord is involved so you are really getting a very nice clear view of what cortical mechanisms can do.
And finally and of great important is that it has a particular sensitivity to the adolescent period. So for example more traditional neuropsychological tasks such as the Wisconsin Card Sort can appear adult-like by 10 years of age, however the task that I will be showing to you has a sensitivity where even adolescents are not performing at adult levels. And I will show what this task is about and I will ask you to look at me for this demonstration. So what we do in the task is the following: we say please look at a center fixation. A light will appear somewhere else in the periphery, the instruction is don’t look at it, look exactly to the opposite location and it should look something like this, very nice and easy. However, what we see when we see mistakes, which we see children committing 50% of the time, teenagers about 20 to 30% of the time, even adults about 10% of the time are unable to do this correctly.
And across psychiatric disorders this is one of the first things to go, and we see the following. Oh shoot, and then look the other way. Now this is very telling, it’s telling us I knew exactly what you wanted me to do because I corrected my movement. But I was unable to engage my executive system and the reflexive system was the one that predominated in choosing the action that finally occurred. So I’ll be presenting studies based on this particular task.
Now the first study that I’m showing you here is a study that we did a long time ago where we used a block design and we presented this task and we compared it to regular eye movements. And we found that a beautiful, elegant very widely distributed circuitry was evident, it was there even in childhood. But there were certain areas where we found developmental differences. One circuitry that we found differences were in areas including the frontal eye fields, the sphere colliculus, the intraparietal sulcus, these are areas that we have found, that have been found in single cell monkey studies to be crucial for the preparatory period. So in monkey studies this circuitry has to reach predetermined level of activity, if not they will fail and the reflexive system will take over. This is not completely there in the adolescent period. When we’ve done behavioral studies where we actually manipulate the amount of time that we provide for response preparation we find that yes, everyone across the age spectrum becomes better with longer preparatory times. But developmentally that difference is still there. So response preparation, response planning, which is actually occurring at the time that you are showing the instruction cue is still immature in adolescents.
We’ve done other studies again to probe what is it about executive function and cognitive control that is immature in adolescents and we did this study and actually I’m showing everybody who has been leading these studies in our laboratory, in this case it’s Katarina Villanova. And we did the study again and when we only consider what is the brain looking like when you are doing a correct trial, which like I said before, children do less, adolescents less, but when we only consider correct trials we see that the brain is actually very similar. And there are only specific regions that are showing differences, including regions in prefrontal cortex. And what we find essentially is that by adolescence these regions are very similar to what adults are doing with those regions. Children are having a more difficult time and they need to engage these regions to a greater degree, which is reflective of greater difficulty in performing this task.
Now as I showed you before there is an increase in the ability to do this task, and this is very important because what is getting better is the rate of correct inhibitory responses. So this is an inhibitory task but similarly in working memory tasks I think it’s very important to understand that the ability to do these tasks in an appropriate manner is available, what is not available is the ability to do this in a consistent flexible and controlled fashion. So what we did is that we were inspired by the attention literature and probed the networks that are known, that have already been mapped, that allow you to sustain an inhibitory set, or sustain an intentional set. And here we start to find very dramatic changes that are occurring through development, so many of the regions are already there by adolescence, of course showing you know greater engagement in childhood. But there is a set of regions depicted in red here, including dorsolateral prefrontal cortex but also parietal regions and visual cortex. This circuitry you need to sustain inhibitory control and this is really only being greatly tapped in adulthood.
Another part that we found was very interesting and still immature in the adolescent period was when we looked at what was the brain doing when you actually did an incorrect response and corrected it. So here you have the results for the adults, so this is the midsagittal section of the brain that has been blown up so we can really look at all aspects of the brain and you see that this area which is the dorsal anterior cingulate cortex showing robust activity when errors are committed. This is an area that has been found is very much involved in monitoring your performance also in allowing you to have a metacognitive view of what you are doing. and to integrate errors so that you don’t continue to commit errors. And we found that that was very much available for this task doing errors, in fact that was replicating findings from MIT where they had done a similar study and found that same area. However that is not really being tapped out in adolescents or in children. And our latest absolute findings looking at longitudinal trajectories of this task is showing that the one thing that very robustly remains is in fact just this, this – the fact that as you get older you are able to integrate the dorsal anterior cingulate cortex.
Now what we looked later, because as I told you the dopamine system and the basal ganglia which includes the ventral striatum are still immature in adolescents. We looked at motivation. We did this task, but this time we said if you do this correctly in some of the trials you can make a monetary incentive, you can make points towards more money. In other trials do the best that you can, you are not going to lose anything, you are not going to win anything. And what we found was very interesting. So first of all here we have by age the reaction time, how quick were you to initiate a correct response. And you can see that regardless of age everybody is a lot quicker when money is involved, not surprising.
Now when you look at the amount of errors that are occurring you can see that adults, they are so good at this and it doesn’t even affect them to have a monetary incentive. They are almost reaching ceiling levels, not quite. But what is very striking is that with younger subjects once you give them the motivation, hm, now they can go something that apparently they were unable to do before. So let’s explore how this actually occurs.
Here is some of the results that we found. I’m showing you the ventral striatum and basal ganglia which is rich in dopamine, as I said before. The results for the left and the right ventral striatum for children, adolescents and adults solid lines indicate rewarded trials, dashed lines are neutral trials, and what we found is that only the adolescents showed a dramatic differentiation in the ventral striatum between incentives and neutral trials. And this is as you can see really driven by the fact that in the presence of incentives all of a sudden neutral trials are really not processed in the ventral striatum anymore. However what is more striking is the fact that when ventral striatum is being very robustly engaged areas of the circuitry that I showed you beforehand that have to do with preparation and being able to do this task correctly are being driven more so in motivation so providing us a view to a possible mechanism as to how they are able to do something that they couldn’t do before. You get the ventral striatum going, you bring up the regions that you need to get the reward receipt, you put them in overdrive so that they can perform this correctly.
We went a little bit further in another study and we looked at the different stages of reward processing. First when you are presented with a cue and you have to actually assess what is the meaning of this cue. Subsequently you look at response preparation and the expectation of a reward. And finally what happens when you get the reward receipt and do the performance. And here again we have very striking results. Black is adults and red are adolescents, and you can see that in the ventral striatum you see that immediately adults were able to engage the striatal activity and then be able to calm it down very – quite quickly. In the orbital frontal cortex you also see that it is only adults that are engaging this region, this is a part of the reward circuitry that processes the more executive aspects of the reward and motivation processing system. By the time you get to preparation there has been a switch, now adolescents are showing a overreactivity of the ventral striatum, again accompanied by regions of the brain that support the ability to gain that reward receipt, so again a mechanism for motivation, bringing up the regions that you need to get the reward in an overdrive type state. By the response time we did not see developmental differences.
So in conclusion for the functional specificity part of the brain with the regional changes that are occurring because of development what we are finding is that underlying limitations and cognitive control in adolescents are immaturities in response preparation, response state, error processing and reward processing. And brain function of cortical and subcortical regions throughout the brain are recruiting by adolescents but they are really engaged differently with development. But I do want to you know emphasize that a lot of it is already there.
Now as I told you before, this whole idea of myelination of interconnectivity has become more and more predominantly important in understanding this later period of development. So we know the regional changes are occurring but we also know that the ability to interconnect to really have the brain working in a collaborative fashion is getting better. We know that there are changes occurring that have to do with Hebbian processing, right, the brain learning that certain connections are working better than other and keeping those, and that this could have crucial changes to behavior that are related to development.
So here are some of the studies that we have done. We’ve looked at the integrity of the white matter using diffusion tensor imaging. And what I’m showing you here is highlighting the parts of the white matter that we found were still immature in adolescence. We did a Voxel-Wise analysis, we used regression methods and we identified what regions are changing throughout the age span. And then we focused on these regions and looked at stage level differences, what is changing from childhood to adolescence, and adolescence to adulthood. And we found that a lot of the white matter skeleton which is represented here in green is already there by adolescence except for association and projection tracks that supports cortical to cortical communication, but importantly cortical to subcortical integration and you’ll see how this develops and how important this might be.
I’m showing you here very recent findings that we’ve done with longitudinal DTI studies and longitudinal studies pretty much always confirm the cross-sectional studies but they provide a way to look at growth trajectories. And what I’m showing you here in all these several, and this is being guided by an MD/PhD student, is basically that the overall growth occurs between 10 and 16 years of age. So from late childhood to adolescence there is a lot of growth that is occurring throughout different pathways and you can see this represented in red throughout different cortical areas.
Late to mature are limbic tracks and these regional terminal zones which is a term that we invented to just really focus on the part of the white matter that is reaching its intended gray matter goal. We found that regional terminal zones that were reaching frontal temporal, parietal and basal ganglia are all maturing until 15 years of age, so a lot is occurring during that time. However last to mature were tracks that included the Uncinate Fasciculus which provide connectivity between the hippocampus, the amygdala, temporal regions, in orbital frontal cortex, and this appears to be continuing into 20 years of age. And we are also finding sex differences indicating that males are a little bit later to mature but at the end of the day reach greater FA values which are reflective of white matter integrity.
Next we’ve been looking at functional connectivity, so why that we are doing that inhibitory task that I explained to you before we looked at causal relationships. And what I’m showing you here, these are actually arrows, we used a process called Grainer Causality where we were able to determine how one region is affecting another region so that we can start to see what are the causal connections in networks that are occurring.
Here I’m showing you what we found, the net – where the causal networks would underlie this performance during childhood and in blue are what was predominant during this age and this is actually replicating some of our earlier findings which is that during childhood where you are doing this pretty poorly you are really relying on posterior regions of the brain. Prefrontal cortex is not really helping you out a lot there. By the time you reach adulthood that basic architecture of prefrontal cortex is there, it is available. And I think this is a very important point, especially when thinking about disorders that might be affecting prefrontal cortex that it is this period that all of sudden you are being asked use prefrontal cortex. So if there is something wrong there this would be a time when you would start to see it. The changes that occur from adolescence to adulthood are an increase in the number of these causal prefrontally guided connections and an increase in the strength of these connections to support cognitive control.
Next I’m going to tell you a little bit about some of the more recent work using resting state connectivity, so this is asking individuals to not engage in a task so that you can look at the spontaneous activity that is occurring in the brain that shows us what areas of the brain like to talk to one another. So it’s a way to probe what is the predisposition of networks in the brain. And we looked at a very particular aspect of this which is called the hub network, and an easy way to think about this is when you think of travel, Atlanta is a network, Chicago, I mean is a hub, Chicago is a hub, Pittsburgh was a hub, no longer a hub, and it’s the area where you are pretty much forced to go before you go to another location. Well the brain seems to be – have an architecture that’s very akin to this idea where you have big areas that take care of a lot of travel and provide a way for different networks to talk to one another. And the way that these areas are found is that you just use every computer resource available to the university and you can proceed to do correlations of spontaneous activity between every voxel of the brain and with every voxel of the brain. And some will arise as having the highest level of correlations, which I’m kind of representing here, and that will determine what a hub is. And in fact this has been done by other people, for example Randy Buckner was one of the initial people to look at this and he found that there was a very robust hub architecture in the brain which he’s, he has looked at in Alzheimer’s disease. I’m not sure people have started to look at this in schizophrenia but I think it would be worthwhile.
So we were interested in seeing what is the hub architecture changing during the adolescent period, during childhood; and honestly we thought that there would be some robust changes. In fact what we found was not that at all. When we looked at this very careful, very carefully and I’m saying we were using extremely high level mathematical processes including graph theory analysis we found that the number of these hubs and the location of these hubs were already there by childhood. The number of connections time degree density was also evident already by childhood. The average strength was also established by childhood, so this complex foundational architecture already there by childhood.
We went even further and we looked at something called betweenness centrality which is a measure of the significance of a hub in supporting a high degree of connection path. This was also already there, not only that this hub network is constructed in the most optimal manner that a network can be constructed, something akin to what you would see in Facebook. So this hub network has some of the properties which are called small world network properties and it’s already there even from childhood. So this is actually extremely striking because it’s telling us that there is a foundational architecture of the brain present very early on that perhaps is providing the basis for where specialization can occur.
Now I think it would be a very important question to ask is things as severe as schizophrenia have an abnormality at this level or in the specialization? So for example by specialization what I mean is how hubs are talking to non-hub areas, right, so how Atlanta is being able to deal with coming to Pittsburgh. And this is where we found the most dramatic changes, which are very similar to what I’ve shown you before, a recurrent theme which is that by from childhood to adolescence you see the integration of prefrontal networks, prefrontal systems being able to talk to other cortical areas, being able to go down and talk to the striatum. It seems that this is coming online in adolescence, nearly mind you. All of a sudden it’s like wow, I have a prefrontal cortex.
And then what is occurring from adolescence to adulthood is actually okay, I have this area now how can I specialize? And in this study what we found were very few specializations, some of them very interesting that had to do with integrating the cerebellum which is not just for drunk driving tests or breathing, the cerebellum really allows a way to have better precision, to integrate learning, to integrate timing, so how to make something that you can do better. And that is what we are seeing is occurring from adolescence to adulthood.
And finally I’m just showing you some of our latest work because we’re really interested not only in understanding why there is so much variability in the adolescent trajectory period but also we want to look at what is the effects of the fact that there might be more dopamine availability during the adolescent period. So here I’m showing that we’re starting to do genetic studies, I’m showing you an example of one gene, the COMT gene, which underlies and has effects in the production of dopamine so that you have these polymorphisms where for example the meta polymorphism ends up with higher production of dopamine and it goes down to the Val/Val which has less production of dopamine.
Now when you are talking about executive function you need the perfect amount of dopamine, too little or too much is not good, you want the exact amount. So you can just start to imagine that the interaction between development and dopamine might be of importance, so we started to look at this. And what I’m showing you here are results first for changes in structural volume in the Caudate and of course we were looking at the frontal striatal system because we are interested in dopamine and we found that yes that in fact even at the level of one gene, polymorphism of one gene we are starting to see that the variability that we see in development can start to be better understood when looking at genetic variability.
We see that you know the Met/Met gene compared to the other ones the polymorphisms have less variability in some ways during adolescence and you can see that there are changes that are occurring from childhood all the way to adulthood. We – people really have only focused on this level, but you can see that there are tremendous changes at the structure and in functional connectivity. These are resting state connections. And we find that look the Met is doing something very different in adolescence than it is in adulthood for connections that have to do with reward processing and connections that have to do with cognition. So I think this is important to understand, also I think that it might have some important implications when we are talking about medications that affect dopamine and how they might have different effects in adolescence and adulthood.
So in conclusion the specific processes that support efficient cognitive control are optimizing in adolescence and they include things like planning, the ability to sustain an executive state to and to engage error and reward processes. Structural and functional connectivity continue to specialize in adolescence supporting frontal top down connectivity which is crucial when you think of what is occurring in mature, in maturation but also what is failing in psychiatric disorders, the ability to have executive areas go down and say no, you are not going to you know react in this particular manner and that the adolescent shift to a better specified and more widely distributed circuitry may be vulnerable to impairment. And as Dr. Lewis presented before we think in a similar manner that the shift itself may trigger an impairment and I’d like to think of this as a stick shift change in a car. I’m sure a lot of you young people may not know what a stick shift is, but the way that I think about this is that you know you are proceeding first gear, second gear, maybe third gear and then in adolescence you are going to go to fourth or fifth gear which now might be using a different part of the engine that might have always been awry or that just that process of going into that gear might be a disturbance that could in fact result in some impairment. So the shift itself could trigger an impairment or the transition and the mode of operation may tap into systems that were already impaired, and I think both of those could be the case.
So finally I want to – it takes a village to do these studies, so I want to thank the people in my laboratory and of course the supports from the National Institutes of Mental Health who although I do basic normative studies recognize the importance of this to understand psychopathology and have always been very generous in their support. Thank you very much.