Skip Navigation
  Print Page

Transcript: NICHD Research Perspectives—February 4, 2013

Skip sharing on social media links
Share this:

Announcer: From the Eunice Kennedy Shriver National Institute of Child Health and Human Development, part of the National Institutes of Health, welcome to another installment of NICHD Research Perspectives. Your host is the director of the NICHD, Dr. Alan Guttmacher.

Dr. Alan Guttmacher: Hello, I’m Alan Guttmacher. Thanks for joining us for another in our monthly series of podcasts from the Eunice Kennedy Shriver National Institute of Child Health and Human Development at the National Institutes of Health. Our guests today are Dr. Bruce Simons-Morton, Dr. Tonja Nansel, and Dr. Ronald Iannotti. We will be talking with them about the prevention research they conduct here at NICHD.

Adolescence is a time when unhealthy behaviors may develop and set the stage for later illness and even premature death in adulthood. For example, many adolescents develop unhealthy eating patterns or begin using alcohol, tobacco, or illegal drugs. As adolescents begin spending more and more time away from home, the influence of peers increases and the influence of family decreases.

The NICHD investigators we will talk with today conduct studies that observe human behaviors or that test new methods to foster health behaviors. This research focuses on the influences of individual characteristics, parents, and peers on adolescent health behavior. Findings from such studies are used to develop strategies for improving or protecting maternal, child, and adolescent health.

Current projects, which today’s guests will talk with us about, focus on three themes: novice teen driving behavior, family management of diabetes, and adolescent problem behavior.

Our first guest is Dr. Bruce Simons-Morton, the chief of the Prevention Research Branch here at NICHD, where all of today’s guests work. He and his colleagues are involved in research on teenage driving, a topic of great concern to parents of adolescents, of course. While young drivers, those 15 to 20 years of age, represent only 6 percent of all drivers, they account for 11 percent of driving-related deaths and 14 percent of all police-reported crashes that result in injuries.

Dr. Simons-Morton first became interested in studying teen driving when his postdoctoral fellow, Dr. Patricia Eitel, proposed that NICHD conduct a study on the topic. They began with a survey of teenagers, asking them about driving privileges and parental supervision of their driving. The researchers were surprised that when most teen drivers got their licenses, their parents gave them the keys with no restrictions at all. This finding led Dr. Simons-Morton to develop a comprehensive research program to identify the risk factors underlying teen driving behaviors.

Bruce, your research has focused on why teenagers are at high risk for vehicular crashes. You’ve done a number of studies exploring driver distraction, the effects of passengers on driving behavior, and what you call gravitational force events—driving tasks such as late braking, rapid starts, and sharp turns. Tell us about your studies and what you have learned about teen driving risk.

Dr. Bruce Simons-Morton: Thanks, Alan. I’ve been studying adolescent health for several decades, and there’s no more important adolescent health problem than motor vehicle crashes, which is the leading cause of injury and death among adolescents. And I’ve been primarily interested in novice teen driving. We know that crash rates are highest early in licensure, and they decline rapidly for about a year and then much more slowly into the twenties. Now American youth get licensed at relatively young ages-- 16 or 17 in most states. In Europe, they get licensed a little later, and many get licensed in their twenties.

So an ingenious study was done to evaluate the effects of young age and inexperience. They looked at crash rates among drivers who got licensed at various ages, and what they found was the same high crash rate right at the beginning of licensure that declined rapidly at first and then more slowly later. But-- the youngest drivers had the highest rate, and they declined more slowly. So clearly this showed an effect of inexperience—no matter what age, crash rates were high—but it also showed that there was something about young age that also contributed to this problem. Now this rapid decline and then more slow decline over time is classic learning curve that you see in every complex behavior—errors when you start a new, complex task. Think about skiing or golf or chess, for example. You’re just terrible, you make a lot of mistakes, and then gradually you get better at that. But the complexity of the task increases, or you challenge yourself more, and so even though you’re getting better, you don’t get to mastery for a long time.

And that’s what we have with driving, which is actually very complex. Driving can be thought of this way, even though young drivers learn to manage the vehicle in a relatively short time. At about 6 hours, most novice teens can back up, they can manage the vehicle. But that’s not the same as driving safely, which requires many frequent complex judgments and a lot of self-control. So what explains the relatively high crash rates of teen novices? We note inexperience in young age—young age associated with immaturity, possibly with risk-taking, and we also know that crashes are higher under certain driving conditions, such as late at night and with teen passengers and certainly with alcohol and drugs.

So we’ve developed a relatively systematic program of research on young drivers, and I’m going to tell you first about the Naturalistic Teenage Driving Study. Until recently, we were limited in our ability to study driving to archival records. You see the records of all crashes, and you can evaluate those, but you know there are some limitations to that. You can also examine each crash and see, you know, what caused that crash but you weren’t—there’s limited information from that. And then you can ask for self-report. So these are all valuable, but it’s not the same, as sort of real time.

And so, recent advances in technology have made it possible to instrument vehicles with complex technology, including accelerometers, cameras, global positioning systems, so you can learn, you can follow drivers and know almost everything about them. This is very new; this is called naturalistic research. And at the NICHD, we conducted the first U.S. naturalistic driving study of teenagers. We recruited 42 newly licensed teens, instrumented their vehicles with accelerometers and cameras, followed the teens and their parents who were driving the same vehicle on basically the same roads for 18 months. And this landmark study provided some very interesting findings. For example, we found that crash rates were very high among the teens and then declined pretty rapidly over the 18 months. But still at the end of 18 months, they were much higher than their parents. The parents have very, very low crash rates, as you can imagine. So, this is consistent with other information from other studies.

You’d be interested in the variability, and we found three trajectory classes. So what this means is over time what was their crash rates? Among the teenagers—like I say, with adults--the trajectory was basically low. Among teens, we found a trajectory of very low crash rates among one group, quite high crash rates or crash and near crash rates among the second group, and then a third group, which is really an interesting group, had high crash rates for about 3 months and then they began to decline down toward the low-risk group. So this is a group that was definitely showing learning. The first group probably was more mature, more self-controlled, something like that. The high group was clearly not learning from their experience, and this middle group was. So that tells us a lot. So that’s crash and near crash, and I should say that a near crash is just like a crash, only at the last moment contact is avoided; and because these studies were small, you can combine crashes and near crashes, and they’ve been shown to be very, very similar.

There’s another important measure of risky driving. We call it kinematic risky driving, it comes from elevated g-force events; so, an elevated gravitational force event is one in which the car, it moves suddenly. This occurs in hard brakes, can occur in very sharp, very fast starts, but usually in hard brakes, sharp turns, and in over-correction maneuvers called yaw. And our study was one of the first to demonstrate that the rate of kinematic risky driving or the elevated g-force of that rate is a very good predictor of the likelihood of a crash or a near crash in a subsequent month. This is information that insurance companies have now really picked up on, and because accelerometer technology is generally available, easy to install in cars, many insurance companies are hoping to have accelerometer data on all drivers so they can charge your insurance according to your risk rate. At any rate, what we did we find about elevated g-force events? Well, not surprisingly, the teens had, on average over 18 months, five times higher rates than their parents. So their parents had an elevated g-force event rate, had these events very infrequently, almost invariably to avoid running a red light. It’s called the dilemma zone: you see the amber turn, you have to make a judgment, and this would often lead to a hard stop, and that would be the elevated g-force event. For teens, however, it was relatively routine: they take sharp corners, they corner, and then they over-correct, they move into traffic and merge and adjust rapidly. So they have five times higher rates than their parents. And, again we studied trajectory classes, you know, they’re not all the same. And we found in this case two classes. We found a group that had relatively lower rates, not as low as adults but lower than the rest of the teens, and another group that was relatively high and relatively invariable; that is, they didn’t change over time. So we have sort of a high-risk and a low-risk group. We looked at a lot of individual characteristics trying to understand: how do these groups vary? And the only individual characteristic that we’ve been able to identify, both for crash and near crash and for elevated g-force events, is the teens’ report of the risky behavior of their five closest friends—either risky driving or other risky behaviors such as substance use. So this is very consistent with like a social influence or social norm that is going on, but more about that in a minute.

We also study secondary task engagement. Teenagers come into driving with texting and phoning habits that adults don’t have. It’s very common for them to use a cell phone to text and so on outside of driving, and so when they start driving the same thing occurs. This is a very unusual study in that we were able, because we have cameras, in every crash or near crash, we would look at the video. And this isn’t real time; this is like a month later when we’re looking, we download the data and we’re looking at it. The coders would, for every crash, they would know: what was the driver doing? So this is the first objective information we have about secondary task engagement. We also looked at adults’ secondary task, and the prevalence varied some, but by and large was higher for the teens than for adults. But what is interesting is we want to know what was the likelihood of a crash or near crash when engaged in a secondary task, compared to the frequency of engaging in that task in a non-crash event. So you sample a bunch of other non-crash events and see how much they’re texting and similar—so you get an odds ratio. And what we found was that among adults, dialing a phone was a risk factor. You’re about three times more likely to have a crash while dialing than not dialing. That was the only risk. In this sample, the adults did not text much, so we didn’t have data there. Teens, however, there was an elevated risk for dialing, texting, reaching for their phone, reaching for other objects, eating, adjusting various instruments on the panel of the car, and drinking—not drinking alcohol but a nonalcoholic beverage.

So, almost everything that they did in the car led to an increased risk. So that’s curious, isn’t it? Why would that be? But it’s sort of consistent with an earlier study that we did. We took another sample of teens and a sample of adults, unrelated, and we had them drive a vehicle, an instrumented vehicle, on a test track. The test track had a signalized intersection, so the driver’s going along, and there is a research assistant with them in the front seat of the car, and actually they got familiar driving the test track and the test track is just a mile long road with a couple of circles and turnouts, but there’s no traffic, so you can do experimentation like this. As the driver was approaching the intersection, the research assistant handed them a cell phone and gave them a task to get some information. And just as the driver would start that task, about 150 feet from the light, we would turn the light amber. So as soon as the driver started on the cell phone task, the light was going to turn amber. What the teens did invariably was they got that information, they were all over that cell phone task; and adults were very clumsy. With the adults, they would dial three numbers and they would look up and see the light is amber, and all of them stopped. With the teens, about 40 percent of them failed to look up and ran the red light. We did that a year later—now they have a year’s driving experience—with exactly the same effect. They hadn’t really learned to divide their attention, whereas adults have.

So there’s a couple of conclusions from this research on secondary task. One is that the type of tasks that are dangerous are those that take your eyes off the forward roadway, and two is that adults don’t like to take their eyes off the forward roadway. They have sort of learned; this is called automaticity. They have developed this innate sense with a lot of years of experience that they want to see what’s going on ahead because if you’re not looking, you cannot react, you can’t avoid an accident, but teens haven’t developed that sort of automaticity.

Now one of the things we thought we would find, but did not, in the Naturalistic Teenage Driving Study was an effect of teen passengers. We did not see an effect of teen passengers on crash rates or on kinematic risky driving; teens drove about equally risky by themselves. Now, when their parents were in the car, they drove in a very non-risky way; they had no crashes and very few elevated g-force events. So we conclude from that that teens know how to drive, novice teens know how to drive in a non-risky way, but they seem to prefer driving in a risky way. But the presence of teen passengers was not associated, which is surprising because the Fatal Accident Reporting System analyses that we have done and others have done find that the presence of a teen passenger increases the likelihood of a fatal crash.

In fact, a study I did with one of my former fellows, Marie Claude Ouimet, analyzing Fatal Accident Reporting System data, we found that not just the age, but the sex of the passenger matters, that male teen passengers had a much higher, fatal crashes were more likely in the presence of male teen passenger, even though either male or female teen passenger increase the risk of a fatal crash among teen passengers. But male teen passengers also increased the likelihood of a crash among adult male drivers, so there’s something strange about teen passengers.

We also did a study where we observed vehicles exiting high school parking lots, and we did that in the Virginia/Maryland area. So we had one observer, observing the vehicles as they exited the parking lot, and they noted who was in the front seat, and were they young or teenager or were they adult because sometimes parents were driving the vehicle. We wanted to distinguish those, and we wanted to know about passengers. And then we had on a road nearby each of these high schools, where those vehicles could get up to speed, we had a camera and a radar gun, and we observed every vehicle. And since we recorded the license number of the vehicles and some other descriptive information, we could identify those vehicles from the high school that were driven by teens and whether they had a teen passenger or not. And what did we find? We could compare the teen drivers with the usual traffic on two measures of risky driving: one was their speed, and the other was how closely they followed the vehicle ahead of them, classic measures of risk.

What did we find? We found that on average teen drivers drove a little faster than usual traffic and a little closer to the vehicle ahead of them on average; but in the presence of a male teen passenger, they drove much faster—about 6 miles per hour on an average 35-mile-an-hour road, which is relatively fast compared to usual traffic, and they drove much closer to the vehicle ahead of them. Curiously, male teen drivers in the presence of a female teen passenger drove just like the usual traffic; that is, their risk was way down. So there is some kind of conservative effect of teen female passengers but a risk effect of male teen passengers. So we have been very curious about the effects of teenage passengers on teenage driving, and our current program of research studies this question.

We wondered: why is there an effect of passengers in the Fatal Accident Reporting System analyses and in our observational study, but not in the Naturalistic Teenage Driving Study? Maybe teenage passenger effects vary by situation such as the driver and the passenger characteristics.

So in the teen passenger studies, we call them, a series of experimental studies testing the effect of teenage passengers on teenage risky driving—and I’m going to explain just the first of these studies; this is called the Teen Passenger Study 1,  just to be creative. And we asked, what is the effect of a risk-averse or risk-accepting teenage passenger on teenage simulated risky driving? So we recruited 80 teenagers who had been licensed for less than a year, and we randomized them to two conditions. And in the first condition, in both conditions, we did a pre-drive priming activity. We wanted them to be deceived by a study confederate, who was a young-looking college freshman actually, who was working for the study, but would act either in a risk-averse or risk-accepting way. And we primed the participants by having them meet just before the drive. And as a risk-accepting passenger, the confederate arrived late and said, “Oh, I’m so sorry I’m late. I went through every yellow light I could. I probably went through some reds to get here, but I just seem to hit every one of them.” But when he came in and assigned to the risk-adverse condition, came in late, he introduced himself by saying, “I’m sorry I was late. Gee, there were so many lights, and I stopped for every one of them.” Priming.

Then they watched a video that showed some very risky driving from the view of—as if you were seeing it from the driver’s perspective, you couldn’t see who was driving, you could just see the vehicle moving through space, you could see this is real road condition. You could see this car changing lanes and speeding up ahead and driving in a pretty risky way. And then we asked the study participant a few questions like, “How risky do you think this was?” and “Would you ride with that person?” And our confederate would respond second, and in the persona of his risk-accepting or risk-averse, so just a little more risk-accepting than the participant.

So now they’re ready to drive in the simulator. And what we’ve done, what we’ve tried to do here is sort of tried to manipulate the perceived social norms of the driver. So now, I haven’t told you what a, you probably know this, but a simulator is a vehicle, but it’s in a lab, and it is surrounded by screens that are connected to the vehicle operations. So, when you step on the accelerator, the vehicle seems to move through space. When you step on the brake, you slow down, and it’s a real world out there—I mean, a seemingly real-world, a simulated world. And in this particular simulation, we had them go through an urban, a rural, and then an urban environment. We told them that they should drive following the rules of the road, and we incentivized them to get to the destination in a certain amount of time. There were a lot of intersections, and we manipulated the amber and red light. So that many times they were in the dilemma zone, and they had to decide to slow down or stop.

Meanwhile, the risk-averse passenger, our confederate, would exercise fairly subtle kind of influence to encourage the driver to follow the rules—or not run red lights. Whereas the risk-averse, risk-accepting passenger rather, says that “I can make that,” and so there was social influence to go faster. So, the participant drove both with and without a passenger so we could sort of control for driving alone, and what do you think we found? Well, there was an effect on risky driving, measured by the amount of time they were in the intersection when the light was red. So this represents that they try to make it through that light. They were pushing the envelope. There was an effect of having a passenger in the car whether they were risk-accepting or risk-averse. So this suggests that drivers, teenage drivers, are kind of susceptible to peers. But there was a significant difference between the risk-accepting and risk-averse passengers. When the participant was with the risk-accepting passenger, they drove in a much more risky way than with the risk-averse passenger. So, what do we conclude? Well, passengers matter, and probably the type of passengers matter.

We were fortunate in this study to collaborate with a young neuroscientist, Emily Falk, at the University of Michigan, whose lab is studying the effects of various psychological tests or tasks on brain imaging, and they were interested in our measure of risky driving. So we decided to do a collaboration, and most of our teens also went to their FMRI lab. I’m going to tell you about one task and the result. So the teens went to the lab and one of the things they did, they did a lot of testing of various sorts, but they were introduced to a game called Cyberball. So this is the classic game in psychology that gets at how people respond to exclusion. So in Cyberball, you’re at a computer, in this case they were introduced to two young people that they told they were going play Cyberball with when they’re in the MRI. And what a Cyberball is, is just basically a ball on the screen, and you can control the ball and send it back and forth between three people. So when it comes to me, I can send it to either of the other two people, and they can send it to either of the two people. So now they have the rules of the game. They are in the MRI; their brains being imaged. They want to know how their brains respond to exclusion. And what happens is, they start passing the ball and they have a screen in the MRI, and they have a little button so they can move the ball. And they are playing Cyberball, everything is going fine, and then gradually the other two start excluding them, and they just pass between each other. Well, for some people, it doesn’t bother them that much. For other people, it bothers them a lot, and this is called costly exclusion. And the brain responds: those for whom exclusion is costly, certain parts of their brain really light up, so you get a nice brain image of it. But what’s interesting about it from our studies is we want to know are those who experience costly exclusion more sensitive to social influence? And indeed, we analyzed the data; a week later they did the simulation. We found that those teens who experienced costly exclusion, they didn’t like being excluded, were more likely to drive in a more risky way with the risk-accepting passenger than with the risk-averse passenger, confirming the hypothesis that there’s variability in how some teens respond to passengers. So we’re continuing this series of studies. Currently we are studying the effective pre-drive mood on the effect of passengers in simulated risky driving. You can see how this would be interesting because it’s a very common type of teenage crash when teens are together and they’re in a positive mood, and then something goes awry.  One way of thinking about this is that when you’re in a very good mood, you’re sort of self-controlled, your guard kind of goes down, and so what we think might be happening when you’re in a really good mood when you start driving is you might be more influenced by either social norms or peer pressure, and that contributes to a crash. Anyway, that’s our next study.

So to kind of wrap up this section, I just want to mention two other things. We have two other naturalistic studies ongoing, one is very similar to the Naturalistic Teenage Driving Study, only we’re recruiting teens—we have recruited teens—when they get their permit. And we are testing the question, what is the effect of supervised practice driving on independent driving performance? So this is a really key question because many driving policies are increasing the amount of practice driving, and we want to know what effect that has. And then finally, in a study you’re going to hear about in a few minutes from Dr. Iannotti, the NEXT study, which follows teenagers from 10th grade up to 5 or 6 years into the future into their out-of-high-school, out-of-home years, we have a sample the we’ve instrumented their vehicles, and we are going to follow them over a long period of time to see how their behavior changes. So, I have been very fortunate to be able to develop a systematic program of research with one study building on the learning from the last, and given the lack of systematic funding sources for young driver and related injury research in the U.S., at least until recently, this may have been only possible as an intramural researcher at the NICHD.

Dr. Guttmacher: Bruce, I know that in addition to examining driving risk and driving behaviors among these new drivers, you’ve also used that knowledge to study ways to prevent teen crashes. In fact, your findings have led you to develop strategies to limit young drivers’ exposure to risk while they are developing their judgment and learning safe driving habits. Can you tell us about these strategies, in particular the Checkpoints Program, which you designed to increase parental management of teen drivers?

Dr. Simons-Morton: I’d be pleased to. The traditional approach to preventing teen crashes has been driver education. Unfortunately, many studies have shown that there are really no safety effects. Driver education works—it facilitates teens learning the rules of the road and getting licensed—but it does not provide any protection. As we discussed earlier, inexperience is a huge risk factor and—so we need—most current approaches aren’t really designed to provide some reduction in risk during the first year or two of driving by limiting driving conditions.

And so over the past two decades, graduated driver licensing has emerged and was gradually adopted in all 50 states. Now what graduated driver licensing is, at least the goal is to reduce exposure to the highest risk driving conditions while novices are learning to drive. So there is a long supervised practice driving period, followed by a provisional licensing period, with limits that vary from state to state on things like late-night driving, some have restrictions on passengers, some have restrictions on cell phones, and so on. So, and then there’s a full unrestricted license usually at the age of 18. Now these policies have been demonstrated to provide safety, they reduce crash rates wherever they’ve been evaluated, but graduated driver licensing varies in strictness by state, and the young driver problem still persists.

So what can be done? One of the things that can be done is, one of the only other ways in fact to limit the teens’ driving once they’re  licensed beyond GDL, is through parental management. We developed a Checkpoints Program, and we’ve evaluated it in a series of randomized trials and demonstrated that it increases parent limit-setting and reduces risky driving and improves driving outcomes. Now basically the Checkpoints Program uses persuasive communications to encourage parents to set limits on the newly licensed teenage drivers by adopting a parent-teen driving agreement. And the persuasive communications are basically norms oriented, suggesting to parents that other parents tend to adopt such limits, that their kids will not hate them if they do, and that it’s really under the control of the teen to gain additional driving privileges by demonstrating that they have been responsible drivers the same way we as parents provided privileges or allowed additional privileges as our teens showed us that they were responsible.

So we’ve evaluated the program in a number of studies and currently Checkpoints is trademarked, and NICHD makes it available to nonprofits such as state motor vehicle administrations and public health departments. We have formed partnerships with the Centers for Disease Control and with the AAA Foundation, and it’s gradually being disseminated in many ways. So while we’ve had effects, we’ve been a little disappointed that the effects are not greater. And so how else can you do this and I’m going to share with you one other study on parental management.

There’s a program called DriveCam now, which uses accelerometers and cameras in cars. The program is largely designed for fleet vehicles. And so truck companies and taxicab companies want to reduce crashes, and they can do that by having accelerometer-based feedback that a manager can observe, and then if the driver is driving erratically, then they’d get a warning. Well, now this program has been adapted for youth. Makes sense, right? Let’s do it with teens. So we wanted to evaluate this program. So we randomized a sample of 90 teens to two groups, and both groups had their vehicles instrumented, and with the DriveCam, it goes right over the rearview mirror, it’s a little tiny box, and on the box is a light. And the light is always green unless they exceed the g-force event, the threshold has been set, and then it blinks red and it tracks what’s happening and what that means is that the footage and only the footage around that event is being saved. Now the camera, the computer is saving the video, is recording all the time, it only saves the few seconds around each event. So, for the first 2 weeks in both groups, we had the DriveCam set on stealth mode so that is there was no light at all. So this is our baseline. And then in group 1, which we called our solo group, we turned the light on so that it was green, and whenever they exceeded an event it went to red—gave them feedback, but only feedback to them, so this is feedback to driver only. In the other group, also we turned the light on, and that group got feedback, but they also were informed that at the end of the week, their parents would get a report of this, and that those recorded video segments for each elevated g-force event would be available for the teen and the parent to watch online. So when we turned the light on, what we found was the group that got feedback and their parents got feedback, their event rates dropped precipitously. The group that got feedback only, their rates actually went up a little bit, probably as the teens were experimenting with turning the light on. And then we’re level, never declined at all. So what we find here in the literature that has evaluated this in fleet operations also says the same thing, only not just the light. The feedback is good management, and here we have more evidence from this study that the feedback to the teen without consequences does not seem to be very important, but the parents have a very essential role to play, and when they are good managers they can reduce this risk.

Dr. Guttmacher: Thank you, Dr. Simons-Morton. That’s a nice, I think, example of the way that studying human behavior really does tell us something about human health and well-being. I think some people’s view of a scientist is someone in a lab with all kinds of beakers and that kind of thing. Certainly at NICHD and NIH we support lots of researchers of that kind of fundamental research, but also the kind of research that you do that really looks at human behavior and, like much of the research NIH supports, it’s about both advancing knowledge, but as you well exemplify using that knowledge to try to actually improve health and well-being. So it’s a very nice, I think, example of the kind of research that we are quite proud of.

Our next guest, Dr. Tonja Nansel, is a senior investigator at NICHD. Like many of the researchers here at NICHD and researchers at other institutions who we fund, she is involved in more than one area of research. First, let’s briefly touch on a topic that affects many youth that she’s been exploring, and that’s bullying.

Tonja, you’ve studied bullying to determine the characteristics of those who are bullied as well as those who do the bullying. I know that your findings show that bullying is not an isolated behavior, but related to other, more violent behaviors. Children who bully are at increased risk for engaging in more serious violent behaviors, such as frequent fighting and carrying a weapon. Perhaps more surprising, your findings showed that victims of bullying also are at increased risk for engaging in these kinds of violent behaviors. Could you tell us more about this research?

Dr. Tonja Nansel: Yes. I’d be happy to. In a nationally representative sample of U.S. children, we found that bullying was in fact a common problem and that kids who bullied were at risk for a variety of adverse health outcomes. Children who bullied typically showed poor school performance, more use of alcohol, they were more likely to smoke; and children who are victims of bullying were more likely to be depressed, to be unhappy at school, and to have fewer friends, not surprisingly. Interestingly enough, we also learned that there was a subgroup of children who were both bullies and victims: so they reported bullying others, and they reported being victims of bullying and that was the group that fared worst of all. They showed essentially all the poor health outcomes of both bullies and victims. But, as you mentioned, what is additionally concerning is that both bullies and victims were more likely to engage in other violent behaviors. In fact, children who were victims of bullying in school were two to three times more likely to carry a weapon or engage in frequent fighting; and bullies were four to nine times more likely to carry a weapon or engage in frequent fighting. These findings challenged some perceptions that bullying is essentially kind of a, well, harmless rite of passage. Prior to this research, granted opinions were mixed in the populace, but a lot of people felt like bullying was really more of a minor issue that children kind of grew out of. So the research that NICHD has done on bullying really raised awareness and raised understanding about this issue; and it resulted in a national anti-bullying campaign as well as numerous efforts by schools across the U.S. to reduce bullying.

Dr. Guttmacher: Thank you, Tonja. I know that another area of research you’ve been involved in is regarding diabetes in adolescents, who typically have difficulty managing their diabetes. That’s in part maybe because hormonal changes that occur during adolescence can affect insulin levels. In addition, many adolescents find it particularly difficult to stick to their daily treatment plan. Moreover, as many parents know, adolescence can be a difficult time—a time when adolescents distance themselves from their parents. This often results in conflict between adolescents and their parents. And as I understand, the purpose of your study is to prevent the care of diabetes from becoming the focal point around which adolescents test their independence from their parents. You’ve recently examined the effectiveness of an intervention to help adolescents manage their diabetes with the participation of their families. Can you tell us what you’ve learned from this?

Dr. Nansel: Yes, it’s been a very interesting and exciting area of research. I’ve always been interested in the challenges youth and their families face in dealing with type 1 diabetes. Years ago, as a nurse at an adolescent psychiatric program. I frequently worked with youth with this disease and saw firsthand how the challenges of disease management intersected with developmental issues in adolescence. But these challenges aren’t limited to youth in crisis. We see challenges in most families who have youth with type 1 diabetes. Most youth with type 1 diabetes have a harder time keeping their disease under control during adolescence. As you mentioned, this is due both to hormonal factors that make the disease harder to control as well as the realities of developmental factors in adolescence.

Any parent of an adolescent is probably familiar with the teen’s desire to have more autonomy, to do things more for themselves and have less parental control. For kids with type 1 diabetes, it’s easy for the disease to become another area of parent-teen conflict; and teens can end up taking worst care of their diabetes because of pulling away from their parents’ assistance. But we found that it’s critical for parents to actually stay involved in a way that helps the teens develop a better ability to manage the day-to-day care of this disease. So, essentially there’s a bit of a transition: early on in the disease; parents are doing more of the disease care for the child, and as the child develops more autonomy and progresses through adolescence, the parent essentially becomes a bit more of a—takes on a consultative role, an assistance role, and their involvement probably has to change from day to day depending on what’s going on with the child and what they need. And then importantly, it’s also important to find ways to help the teens better take on, as they do take on by necessity more and more, responsibility for their disease, we need to find ways to help them do that more effectively so that their disease control doesn’t suffer.

So we developed and tested a clinic-integrated behavioral intervention designed to help families do just that. Now families with type 1 diabetes receive clinical care usually about once every 3 to 4 months, and, as you’re probably aware, managing type 1 diabetes is very intensive: it involves checking blood sugar multiple times a day, often five or six times a day, responding to those blood glucose fluctuations, accounting for eating, exercise, stress, sleep, etc. At each meal, children with type 1 diabetes typically have to estimate how many carbohydrates they eat and then calculate their insulin dose based on that. So it’s essentially another job on top of the life they already have. So the approach we developed was designed to help them better address the challenges of this very intensive disease. It’s centered around a structured problem-solving method.

In this approach, families talk about what area of diabetes management they have difficulty with, and then they jointly set a goal for improving care in that area. So, for example, families might—perhaps they’re currently checking blood sugar four times a day, and they should really do it five or six times a day—so they may set a goal to increase the frequency of blood glucose monitoring. Or perhaps they have difficulty estimating carbohydrates’ content, particularly after eating out or away from home, so they might set a goal for improving the accuracy of the carbohydrate counting. Those are just a couple of examples of the kinds of goals they set. Now this goal was actually set by the parent and child talking together with the assistant, who we call the health advisor, the person who guided them through this process. And there might be cases where the child wanted to work on one thing and the parent thought they needed to work on another, and we really worked to help them jointly set this goal so that it was something that both people were invested in and that the child had a sense of ownership over what they were doing. So this way, families collaborated together, and teens could have an input on how their parent could be the most helpful. It’s much easier to accept help from someone if you’ve asked them for that help, rather than feeling like it’s being imposed upon you.

In our study, the approach was tested in multiple clinics across the United States, and the adolescents of families who took part in the program had significantly better glycemic control than families in standard care. This program was implemented across 2 years of care. So each time when the child had a health care visit, they also went through this problem-solving process. And what we saw in the children who were not part of the program, the usual care group, we just saw the typical deterioration in glycemic control that happens during adolescence. And in the children and their families receiving this program, we saw that for the first little bit in the study, they followed that same path but after they had a couple of these visits with the health advisor using problem-solving method, they actually had a turn in their level of blood glucose control, and then we saw improvement over the last part of the last half of the study. These were very encouraging findings because this has actually been the first behavioral program to show improvement in glycemic control among this, among adolescents, which is such a difficult time in this disease. Importantly, families like the program, they like the approach.  We had one mother say that she really liked this new method of care when they were assisted to identify problems and find ways to solve them. So this approach could really be a way to improve the way that health care is delivered.

Dr. Guttmacher: That’s very interesting, Tonja, I know you’ve also been examining another aspect of life with diabetes that could be a challenge for children and adolescents, and that’s diet. Most adolescent diets are low in fruits, vegetables, and whole grains and high in saturated fat. And, of course, poor diet quality is likely to persist into adulthood, raising the risk of long-term medical consequences for anyone, particularly those with diabetes, so achieving and maintaining healthful dietary change, while a challenge for all of us, is particularly important for those with diabetes. Can you tell us what you’ve learned about the diets of children and adolescents with type 1 diabetes?

Dr. Nansel: Yes, I’d be happy to. This has just been a fascinating area. As we were working with families with type 1 diabetes, we found that this was a reoccurring theme—that both children and their parents were concerned about eating more healthfully and diet has really been a somewhat neglected area of research in type 1 diabetes. There’s been necessarily so much emphasis on the many advances we’ve had in treatment regimens that diet has taken a bit of a backseat. As you know, Americans in general don’t eat as healthfully as we should. And for youth with type 1 diabetes, eating healthfully can help with keeping the disease under control and is important for preventing long-term complications such as cardiovascular disease.

In our research, we found that youth with type 1 diabetes are eating just as poorly as U.S. youth overall and, in some ways, possibly worse; there is evidence that they eat less fruit and more saturated fat. We examined the diets of youth with type 1 diabetes and found that almost half of the calories in their diet came from processed grains, chips, and desserts. That doesn’t leave much room for getting enough healthy food. We also found that kids who ate poor-quality diets were more likely to be overweight, even after controlling for the amount of calories they consumed.

On the encouraging side, we found that some of the perceived barriers to eating healthy may not be as strong as people may think or as we had thought. A lot of families were concerned that eating healthfully might cost too much; and in fact when we looked at the cost the family spent on their food, the people who were eating more healthfully weren’t spending any more than the families who were eating less healthfully. We simply saw that there was a shift in cost—that the families eating less healthfully were spending more money on chips, sweets, packaged foods, etc., whereas the families who ate more healthfully were spending more money on fruits and vegetables—but the sum total was the same. So that was very encouraging because we want to make sure the interventions we develop are applicable for families across a range of situations.

We also directly tested the effect of diet on blood sugar control in another study using a device called continuous blood glucose monitoring. This device measures the person’s glucose levels every 5 minutes while it’s worn; there’s a little catheter that goes under the skin, it records the glucose level—the tissue, which then corresponds to blood glucose. And this allowed us very precisely see what dietary manipulation would do. So we brought these kids into a control setting and fed them two different diets on different days. One of the diets was a healthy version of the standard American diet; it met all the American Diabetes Association guidelines and would be perceived as a pretty good diet for a kid. On the other day, we fed them a minimally processed diet, a lower glycemic index diet, consisting of foods like brown rice, fruit, etc.—all foods the kids were actually quite willing to eat. And we compared their glucose levels on the two different days, and we found a very striking difference. Kids had much better blood glucose levels when eating the less processed diet; and, interestingly enough, we repeated this not just in a controlled setting, but in their home setting with 1 day where they ate just whatever they would normally eat and on another day where they followed instructions to eat a less processed lower glycemic index diet. The findings were identical, so families were able to do this at home.

So currently we’re testing the effect of the behavioral intervention designed to help the whole family eat more healthfully. As you know, it’s important that diet is something that is done by the entire family. We would not want to just isolate the child with type 1 diabetes and say, “You need to eat healthfully, and your brothers and sisters and mom and dad don’t.” Obviously, there are benefits to the entire family, and it’s going to be much more successful if the entire family eats more healthfully. In this program, children and their parents set goals for increasing their intake of fruits, vegetables, whole grains, legumes, nuts, and seeds in each meal of the day. The program integrates the applied problem-solving process that we used successfully in a previous study as well as activity-based education and an interaction style called motivational interviewing, which is designed to help people really build their own internal motivation for wanting to engage in a behavior change. We specifically designed the program to help families find easy and affordable ways to eat healthy. So we’re looking forward to seeing how children’s diets improve as a result and how that impacts the diabetes management.

Dr. Guttmacher: Thanks very much, Dr. Nansel. Our next guest, Dr. Ronald Iannotti, is a staff scientist at NICHD who is investigating adolescent behaviors associated with obesity.
According to the Centers for Disease Control and Prevention, more than one-third of Americans between 2 and 19 years of age are overweight or obese. These young people are at increased risk for asthma, diabetes, painful joints, high blood pressure, and other conditions that can deeply affect health. They also are more likely to become overweight or obese adults and to face serious health problems as they age. And preventing adolescent obesity has proven difficult. To a great extent, this is due to the challenge of changing people’s eating and activity patterns, which is difficult at any age.
Dr. Iannotti is the principal investigator for two important studies of adolescent health behavior: The Health Behavior in School-Aged Children project and the NEXT study. We spoke briefly about the challenges in designing the NEXT study in our November podcast. Today, Dr. Iannotti will tell us about the study’s work to identify the beginnings of adolescent behavior patterns that later lead to obesity and other adult health risks.

Ron, can you tell us a little more about the design of these studies, and what you’ve learned so far?

Dr. Ronald Iannotti: Yes, sure. Thanks, Alan, and I’d like to thank you for inviting the members of the Prevention Research Branch to present here. First, let me talk about the Health Behavior in School-Aged Children study. It’s actually an international collaboration involving the U.S., Canada, and almost all of the European countries. It’s a quadrennial survey of adolescents 11, 13, and 15 years of age. In the U.S., we actually sampled a nationally representative sample of students in grades 6 through 10, and it’s actually a broad survey. The data from HBSC, for example, contributed to the work that Tonja was talking about with bullying. We also look at substance use and a number of other health behaviors, but what I’d like to talk about is what we call obesogenic behaviors and those behaviors that have been shown to contribute to the obesity epidemic in adolescents. They include physical activity and sedentary behavior; sedentary behavior is sometimes seen as the absence of physical activity, but it really has some independent effects on health. It also includes diet. As Tonja mentioned, we really recommend that adolescents get a diet that’s high in fruit and vegetable intake and low in things like french fries, snacks, and sugar, foods that are high in sugar like sweetened soft drinks and candy.

And the other thing about HBSC is it’s conducted every 4 years, as I mentioned, and the U.S. has been involved in the study since 1997. And one of the studies we’ve done, we looked at the trends and behavior during the last decade from 2001 to 2009. We have both good news and bad. When we look at body mass index, which is an indicator of obesity, we find that, we know that it’s been increasing over the last four or five decades. We find in the last decade that it may be leveling off, and this is good news, so instead of continuing to increase in adolescence, it may be flattening that curve or turning that trend. One explanation for this would be looking at those behaviors, and we found that physical activity actually increased during this period, and sedentary behavior actually decreased during this period, particularly television viewing. Unfortunately, other sedentary behaviors like computer use, game playing have increased during this time, more than making up for the time, the decrease in television viewing. We also looked at diet, we looked at fruit intake, vegetable intake, as well as candy, sweetened soft drinks, chips, and fries, and those also have decreased over this period, although slightly. So, it appears as though some of the public health efforts to change the health behaviors of adolescents may be succeeding: physical activity is increasing, TV watching is decreasing, and diet is generally getting better. This probably contributes to this bending of the curve. Unfortunately, as I mentioned, computer use and game playing has increased during this time, and so that may be partially contributing to the fact that we are not seeing a decrease in obesity during this period.

One of the things we did was to look to see if there are patterns in adolescents’ behaviors in terms of these obesogenic behaviors. We found that about a fourth, or one in four adolescents, engage in what we would call a helpful pattern. These are adolescents who, a majority of these adolescents meet the guidelines for getting at least 5 hours of physical activity a week and less than 2 hours of television viewing a day. They also were higher than other adolescents in terms of fruit and vegetable consumption, but they still didn’t meet the recommendations for five servings a day. They’re also low in terms of consumption of sweets, sweetened soft drinks, chips, and fries. So we call this the healthy pattern. Unfortunately, three out of four kids did not exhibit this pattern. About a fourth, or one in four, of the children engaged in what was clearly an unhealthy pattern. They were kind of the antithesis of the healthy kids. They were low in physical activity, high in sedentary behavior, low in fruit and vegetable consumption, and then high in consumption of those foods that we identified as risky foods; and then, about two in four, or 50 percent, of the kids—but which we’re calling the typical adolescent—have a pattern that’s kind of in between the two, but they were really noted for low consumption of all the foods: they reported low consumption of fruit and vegetables and low consumption of the risky behavior foods.

So it’s interesting then that there is plenty of room for improvement in the U.S. adolescent in terms of these obesogenic behaviors. Now one concern would be, how do you change these behaviors or what influences these behaviors? Being an international study, we were able to compare motivations for physical activity across the U.S. and Europe, and we identified three categories of motivations. One is the healthy motivation—basically the message you usually give in schools, that you want to engage in these physical activities to improve your health, to control your weight, and then we saw social motivations. Kids reported that one of the reasons they engage in physical activity is so that they can spend more time with friends and meet new friends. And then the third area we would call achievement or more externally oriented motivations, and those are to do well in sports, to win, and to make their parents happy. We found that the social motivations and achievement motivations were high in all the countries and were positively correlated with physical activity.

Interestingly, health motivation was inconsistent across when we compare the U.S. to Western European countries and Eastern European countries. In the U.S., health motivation was unrelated to level of physical activity in our adolescents. In Western Europe, it was positively related; and in Eastern Europe, where they really don’t have much of an obesity problem, it was negative related. So kids who reported that the reasons they engage in physical activity was to improve their health actually were less likely to get physical activity in Eastern Europe. Another influence on these health behaviors is body image, and we usually think that if a child’s obese, they perceive themselves as obese or overweight, they perceive themselves as overweight. But in fact, many adolescents think of themselves as overweight or obese when in fact they are not so, even if you control for their actual weight status using body mass index, body image seems to be more important in predicting these behaviors; and kids who have poor body image are actually less likely to engage in physical activity, more likely to engage in sedentary behaviors, and more likely to engage in other risk behaviors—for example, smoking. So if we were going to intervene to improve these health behaviors, we might focus on something like how the adolescents perceive themselves, how they perceive their weight status.

We also looked at, you know, being a national study, we could look at schools and school environments and how they may influence these behaviors.  So turning to diet, we ask the school administration to provide information about the school environment—things like access to areas to play, but also we asked about vending machines. And interestingly we found that of those schools that have vending machines, there is a direct relationship between the kinds of foods offered in these vending machines and the diet of adolescents at these schools. So there were school differences depending on what they offered in their vending machines. And these effects were significant for the younger grades—in our case, sixth, seventh, and eighth graders—but didn’t seem to have an effect on the older students in ninth and tenth grade. And then more at home, we looked at TV viewing and how that might influence diet, and we found that kids who watch more TV are more likely to consume more of the unhealthy foods, they’re more likely to have lower intake of fruits and vegetables and higher intake of sweetened beverages and sweet snacks as well as the fries and chips.

Interestingly, we also asked kids whether they actually snacked while watching TV, and those kids also are more likely to snack on unhealthy foods, but they also reported they are more likely than kids who didn’t snack when watching TV to consume fruits. So this might offer an opportunity for intervention if parents make sure they have available, when kids are watching TV and snacking, fruits. You can cut up apples, you can section oranges, and that may be a natural pathway for kids to improve their diet, if they are going to snack while watching television. Now with regard to the other study that we’re doing, one of the disadvantages of the Health Behavior in School-Aged Children study is that it’s cross-sectional. It’s a different sample at every wave that we study them, so we can’t really look at any kind of causal relationship between the behaviors and overweight.

To address this, starting in 2010, we’ve been studying a national cohort of U.S. students in 10th grade, starting when they’re in tenth grade; we’re following them for 6 additional years until the years after high school and really through those years when they’re starting new careers or attending college or establishing a family. So this will be a really interesting study because we’ll be looking at how these behaviors during high school may predict behaviors after high school. Plus we can see transitions in establishing a family and starting a job and going to college and how these transitions may affect these behaviors. Now we’re just beginning to look at these data.

Paul Albert, a month ago, talked about the sleep data that we’re gathering; and so in addition to physical activity, diet, we are also looking at sleep, which has been shown that it may be related to obesity. And in his work with these data, we found that for adolescent girls if they were overweight or obese the pattern of activity is different during the day. Not only do they engage in less physical activity, but there’s a shift in the cycle of sleep and wake time in that the girls who were overweight or obese actually begin their activity later in the day and then stay up longer at the end of the day, and these differences in sleep patterns could be related to the obesity. In addition, to surveying the kids annually, we have a subsample of kids—half of them are overweight or obese, and half are normal weight—which we are studying more intensively. We’re having them wear accelerometers so that we can measure their physical activity. We are having them wear a sleep watch so we can monitor their sleep patterns. We’re having them report on their diets to 24-hour recall three times. We also have collected genetic material from them. In the first wave in 2010, we measured blood pressure and we did analyses of their blood so that we can look at their blood cholesterol levels, you know lipids. We will look at uric acid, cotinine, as well as C-reactive protein, which indicate potential inflammations and risk for heart disease. And from these data, we are looking at the relationship between measured physical activity rather than self-reported physical activity and some of these outcomes.

And we found in our preliminary work that physical activity seems to be related to diastolic blood pressure in these kids and that this relationship is moderated by their weight status.  That is, physical activity has a greater effect, or a positive effect, in reducing diastolic blood pressure in kids who are overweight. So this again suggests that potential interventions because increasing physical activity will have a more substantial effect for kids who are overweight. Currently we are gathering these data again; we are actually collecting blood and measuring blood pressure in this sample of children. We’ve had very good retention: it’s about over 90 percent of the kids have stayed with the study for 4 years. And so we’ll be able to look more longitudinally now at kids who’ve been engaging in physical activity over the previous 3 years and how that may relate to these cardiovascular outcomes.

Dr. Guttmacher: That’s great. Thank you, Ron. That brings us to the end of our podcast for this month. I’d like to thank Dr. Bruce Simons-Morton, Dr. Tonja Nansel, and Dr. Ronald Iannotti for joining us today and for sharing with us some of their research on behavior and how it has an effect—lots of effects really—on adolescent health. I’d also like to thank our podcast listeners for joining us and for your interest in our work at NICHD.

For more information on any of today’s topics and many related topics, visit www.nichd.nih.gov. That’s www.nichd.nih.gov.  

I’m Alan Guttmacher, and I hope you will join us for more NICHD podcasts as we post them on our website each month.

Announcer: This has been NICHD Research Perspectives, a monthly podcast series hosted by Dr. Alan Guttmacher. To listen to previous installments, visit nichd.nih.gov/researchperspectives. If you have any questions or comments, please email NICHDInformationResourceCenter@mail.nih.gov.

 

Back to Research Perspectives.

Last Updated Date: 02/21/2013
Last Reviewed Date: 02/21/2013
Vision National Institutes of Health Home BOND National Institues of Health Home Home Storz Lab: Section on Environmental Gene Regulation Home Machner Lab: Unit on Microbial Pathogenesis Home Division of Intramural Population Health Research Home Bonifacino Lab: Section on Intracellular Protein Trafficking Home Lilly Lab: Section on Gamete Development Home Lippincott-Schwartz Lab: Section on Organelle Biology