Teenaged Brain: Part 1

July 28, 2009
John J. Medina, PhD
Volume 26, Issue 8

This statistic is as familiar as it is startling. According to the National Comorbidity Survey-Replication (NCS-R), the peak age of onset for any disease involving mental health is 14 years. True for bipolar disorder. True for anxiety. True for schizophrenia and substance abuse and eating disorders. The data suggest that most mental health challenges emerge during adolescence. If true, this brings to mind an important developmental question:

This statistic is as familiar as it is startling. According to the National Comorbidity Survey-Replication (NCS-R), the peak age of onset for any disease involving mental health is 14 years. True for bipolar disorder. True for anxiety. True for schizophrenia and substance abuse and eating disorders. The data suggest that most mental health challenges emerge during adolescence. If true, this brings to mind an important developmental question:

What is up with adolescence?

The teenaged years are chock-full of byzantine, intricately timed, molecular processes that have to be closely choreographed and deployed in a specific sequence to accomplish their sexual mission. Do these extraordinarily complex developmental processes go awry in some children as they evolve through adolescence? Do these changes create, or at least contribute, to future mental disorders? Is this one way to get at what are sometimes called genetic “trapdoors”-DNA-based psychopathologies that do not show up until a certain developmental milestone is reached?

These are important issues. Most of the mental health challenges that emerge during puberty have real staying power. The symptoms tend to be more severe. Many go undetected in the early formative stages of the illness and comorbid disorders often develop. These complications can create problems in determining the correct diagnosis, and make it difficult for the clinician to select the treatment strategies with the greatest probability of success.

Researchers face similar daunting challenges in attempting to understand the cellular and molecular basis for such disorders. Fortunately, fairly recent findings have provided a ray of hope-potential illumination for both clinician and scientist. From gene to cell, we are beginning to learn more and more about the neurobiological maturation of the brain transiting through adolescence. The question is, Does any of this knowledge help us understand the NCS-R data?

In this column and the next, we will explore the developmental biology of the so-called teenaged brain, focusing first on cellular studies, then on behavioral ones. In this first installment, we will address specific aspects of the brain’s developmental trajectory. We’ll look initially at structural changes and then focus on the notion of the canonical “teenage brain” behaviors.

In the second installment, we will discuss how changes in these developmental processes may contribute to the emergence of mental disorders.

Alterations in Brain Structure

The past 20 years have witnessed extraordinary progress in our understanding of the functional and structural development of the human brain. Some of the best stuff has been obtained using noninvasive brain imaging technologies, such as fMRI. Though fraught with the potential for misinterpretation (please see my 2 articles online, http://www.psychiatrictimes.com/display/article/10168/1401497 and http://www.psychiatrictimes.com/display/article/10168/1425694?pageNumber=1), even the most conservative interpretations reveal a fantastic world of molecular and cellular activities during development. Efforts examining how the volume of gray matter changes over time has been particularly fruitful, and is explained below.

Initial evidence demonstrating volumetric gray matter changes focused on 2 temporal processes. The first efforts appeared to detect a gradual increase in gray matter volume during the elementary years that reached its peak at puberty. The second effort showed that the volumes appeared to gradually decline after this peak. The “final” adult form of the brain was fully formed by the time students were ready for college.

The problem with this standard model is that it was oversimplified. Closer inspection of brains developing in the elementary years revealed a much more complex picture. Volumetric increases of gray matter followed by a tailing off only occurred for certain cortical regions, such as the frontal and parietal lobes. The temporal lobes didn’t follow this trajectory at all. Actually, the volume of some cortical regions (such as the superior temporal gyrus) was shown to decline during the elementary years.

The picture that emerged was far more complex than previously supposed. It is now apparent that different parts of the preadolescent brain undergo developmental changes at different times-and at surprisingly different rates.

Research continued into the second temporal process: What changes, if any, occurred in the postpuberty brain? Once again a more complex picture was revealed. Some of the previously observed findings of volumetric decline were confirmed in these studies. The postcentral gyrus, for example, underwent a fairly rapid increase in volume by age 10; it then leveled out for a time and then declined fairly steeply until around age 20. The mid-dorsolateral frontal cortex monitored during the same period followed exactly the opposite pattern. Its volume actually started declining around age 10, bottomed out for a while, then began slowly to rise until around age 20.

At the same time gray matter was being examined, other researchers were investigating white matter volumes. Just to make things more complicated, a developmental pattern very different from the gray matter trajectory was observed. White matter volumes were found to undergo a clearly linear increase beginning during childhood. This increase was not arrested at puberty, however, nor did it even slow down significantly. Working something like the Energizer Bunny, white matter volume kept rising and rising throughout young adulthood until around age 30-depending on your gender. The slope of these age-related changes was steeper in males than in females.

Do these changes mean anything in terms of adolescent behavior? The answer is clearly no-or at least, not yet. From a cellular/molecular point of view, however, they suggest a great deal. Alterations in cortical volume are sometimes associated with the changes in synaptic pruning, for example, usually couched as density measures. (It’s a familiar developmental idea: synaptic connections that are overproduced during childhood need to be pruned to achieve their final mature state.) Do these changes in cortical volume indicate changes in synaptic pruning during adolescence?

The answer, interestingly enough, is “No.” There certainly are changes in synapse number in a variety of cortical regions during adolescence. But clear data now exist which show that the decrease in synaptic density is insufficient to account for the observed dramatic changes in volume.

What then accounts for these changes? Although speculation abounds, the real answer is that we just don’t know. One interesting explanation attempted to take into account the simultaneity of the developmental patterns of both gray and white matter. Because white matter usually indicates an increase in the myelination of nerve fibers, it was thought that a processive myelination pattern was being observed. (As you recall, myelin is the fatty sheathing that provides an electrical insulating function for the conduction of neural signals. An unmeylinated neuron can conduct a signal at about 1 meter per second. A myelinated one can conduct an impulse at about 100 meters per second).

Were myelination patterns really being observed? Researchers took this into account and then combined the idea with the gray matter volume data. It was possible that less gray matter was being observed simply because these neural tissues were in the process of being “whitewashed” (myelinated). Covering over gray matter with white would explain the loss of gray matter volume while simultaneously explaining an increase in the white. There is some empirical support for this myelinating notion, particularly with intracortical fibers, which are processively myelinated throughout brain development in the early years.

It turns out that this explanation, satisfying as it may be, does not tell the whole story. Data exist that challenge the notion that myelin has any involvement in the observed developmental changes. There are other ways to measure myelin than just looking for white matter (using a technology whose metric is termed the MTR [magnetization transfer ratio] index). When these technologies are deployed, one sees there is actually a decrease in the amount of myelin per unit volume in the developing brain-even while the amount of white matter is increasing. Sound contradictory? It is the reason I said “we don’t know” in attempting to hypothesize relevant mechanisms.

Regardless of our ignorance, it is clear that understanding puberty-associated changes in brain structure is not as simple as saying “lots of things happen to it at puberty.” To be sure, lots of things are happening to brain structure during adolescence, but changes are observed throughout the brain’s developmental narrative. Indeed, the final chapter in white matter growth does not end until the third decade of life, giving the human brain an odd trophy: it is the last organ to finish its postnatal developmental program.

Functional Changes

What about behavior? It is one thing to observe architectural changes in the brain. It is another matter to say that such changes are responsible for the canonical behaviors associated with adolescence. Are any of these structural alterations in the brain associated with behavioral changes as well? That’s a tough question, and the answer is mostly incomplete.

Things have not been helped by the creation of what some in the popular press euphemistically call the “teenage brain.” You may have heard of this near-mythological idea, which has been mostly an attempt to explain the erratic behavior of many adolescents. It is usually couched in terms of the developmental regulation of executive function. As you may recall, executive function is a suite of behaviors involved in such disparate activities as planning, foresight, impulse control (actually, inhibition), and aspects of working memory (Figure). Risk-taking and sensation-seeking behavior, especially among males, has also been mentioned as being part of the story of the teenaged brain-behaviors adolescents are known for having in short supply. From a neurological perspective, the story usually goes something like this:

• Executive function is mediated by areas in the prefrontal cortex.

• These areas are not functionally connected to other regions of the brain that are involved in emotional regulation, risk taking, or sensation seeking. (The limbic system and the nucleus accumbens are the usual suspects.)

• The regions involved in emotional reactions do not begin to form functional connections with regions mediating executive function until the child is well into adolescence.

• We should therefore understand that certain types of “foolish” adolescent behavior occurs as a result of immature or non-existent connections in these brain structures.

How much of this actually holds up to scrutiny?

Fortunately, there are many aspects of this story that are falsifiable, and researchers have gone after it with robust energy. The findings are actually quite surprising. Some parts of the myth hold up well, while other parts don’t hold up at all. Here are the salient features:

First, it is not true that behaviors associated with executive function are unavailable to teenagers. Many executive functions are developed in the elementary years. It is an uneven transit, however: some kids show great executive maturity given their chronological age, and others show less skill.

It is true that executive function can be dramatically improved as children evolve through puberty, which may be part of the myth’s origin. The most time-sensitive behaviors are revealed in delayed gratification tests and in tests of planning and foresight tasks. Behaviorally, preadolescents don’t score very high on such tests. After the fires of puberty die down, however, they often score much higher.

These improvements correlate nicely with noninvasive imaging observations. Many fMRI images show dramatic increases in blood flow between prefrontal and parietal cortices as these children age (assayed under a variety of executive-testing conditions). These changes hold up, even when individual differences in brain structure are taken into account.

Behaviors like teenaged predilections toward sensation and risk-taking do not hold up as well, however. The idea is that regions in the brain naturally involved in reward-seeking should be more actively engaged in teenagers under tests that require reward seeking. They should then die down as the predilection for the behavior diminishes with age. The focus is usually on the nucleus accumbens and even the ventral tegmental areas, which makes sense, given their natural role in reward-seeking behavior.

Not everybody in the teenaged test cohorts shows such dramatic increases, however. Some researchers have actually found the opposite-a decrease in blood flow associated with increased risk-taking behavior. Some studies show increases to these areas in response to reward-seeking tasks, regardless of age. There are also the normal “chicken and egg” questions that plague so much of this type of research. Does the increase in blood flow create the behavior, or does the behavior create the increase in blood flow?

The Next Installment

As can be seen by looking at just 1 slice of neuroanatomical research and 1slice of behavioral research, much more work needs to be done before a complete explanation of teenaged brain development is in our grasp. And we have yet to address the central focus of this series: the relationship between whatever changes we agree actually exist and psychopathologies originating in adolescence. From the neuroanatomical or behavioral work mentioned above, what could possibly explain the data from the NCS-R?

Our next column is designed to address exactly these issues. There is substantial reason to believe that adolescent-based emerging psychopathologies are the result of interactions between 2 factors: anomalous adolescent maturation processes and psychosocial experience. It is as familiar as your latest ruminations on nature versus nurture put together this time in an attempt to explain a mental health statistic that is as familiar as it is startling.