Imaging of Cognitive Impairment from Soccer Heading-Related Brain Injury

Study indicates there may be a heading threshold above which the risk for brain damage increases significantly

by Jim Schnabel

June 29, 2015

An Interview with Michael L. Lipton, M.D., Ph.D., F.A.C.R.

Michael Lipton headshot  Professor of Radiology, Psychiatry and Behavioral Sciences
Associate Director, Gruss Magnetic Resonance Research Center
Albert Einstein College of Medicine
Director of MRI Services
Montefiore Medical Center

Dana Foundation Grantee, 2012-2015


Dr. Lipton pioneered the use of MRI technology to detect mild traumatic brain injuries (mTBI) from concussions. Such injuries, which may bring cognitive and behavioral impairment and even neurodegeneration later in life in some individuals, is increasingly seen as a major public health problem–in particular for those who play contact sports. According to the US Centers for Disease Control: “From 2001 to 2009, the number of annual TBI-related [emergency room] visits increased significantly, from 153,375 to 248,418, with the highest rates among males aged 10–19 years.”

With support from the Dana Foundation and the National Institutes of Health, Lipton has conducted neuroimaging studies on soccer players, who frequently jolt their brains by “heading” the ball. His goal is to understand better the relationship between head impacts and resulting brain damage and cognitive impairments.

Dana Foundation: When did you begin using MRI-based methods to evaluate mild traumatic brain injuries?

Michael L. Lipton: Clinical referrals of patients with persistent postconcussive symptoms and cognitive and behavioral impairment related to TBI have long presented a difficult problem because standard neuroimaging is not very revealing. In the early 2000s, however, a new MRI-based technique, diffusion tensor imaging (DTI), was just becoming available, which by its nature seemed like an appropriate tool to detect the injury to CNS axons that occurs in TBI. I decided to look at how it might be applied in the clinic to get some traction on this problem.

DF: How do you use DTI to detect concussion-related brain damage?

MLL: DTI allows us to characterize the direction of movement of water molecules in tissue. In particular, we can look at a measure called fractional anisotropy (FA), which indicates how uniformly or coherently water molecules move in a given brain area. In normal brain “white matter,” which is composed essentially of bundles of tens of billions of nerve fibers, or axons, water tends to diffuse predominantly along the lengths of the axons. You could think of these axons as tubes or straws, with water able to run along or within, but not across the walls of the tubes. This movement predominantly in one direction is termed anisotropic diffusion. In the normal case we detect a high degree of this diffusion anisotropy, which we measure as high FA.

In concussion or mTBI, the underlying pathology that is the source of symptoms and problems for the patients is called traumatic axonal injury. That doesn’t mean that at the moment someone is hit in the head, their axons are immediately torn asunder. This initial injury sets off cellular and molecular processes that in some people, over time, can lead to a substantial degree of axonal degeneration and loss of function. In a brain area with this injury, the usual barriers to water diffusion—the membranes and myelin sheaths of axons (those tubes I mentioned)—will be disrupted, and so the direction of water diffusion measured by DTI will be less uniform, more random. Thus the FA values will go down.

So we detect low FA as a correlate of the change in the brain’s microstructure due to injury.

Michael Lipton feature content
Dr. Lipton has helped pioneer the use of DTI to evaluate head injuries.
Photo courtesy of Albert Einstein College of Medicine.  

 DF: Why did you decide to study soccer players?

MLL: In my clinical practice, I saw people who came to the emergency room or a doctor’s office with a concussion or head injury. But a research problem we face is that when you examine a patient after an injury, and you identify something that looks abnormal, you can never fully characterize the relationship between the head injury and the changes in the brain caused by the injury, because you cannot quantify the magnitude of the injury and don’t know how the brain looked before it.

I decided that soccer players were a potential model system for studying mTBI, because, first, we know in advance that they’re going to have mild impacts to the head. So we can bring them in and look at them, let them go out and play, and then look at them again. We can ask them how many headings they remember doing and the kinds of symptoms that they’ve had or haven’t had. We can then try to relate that level of exposure—the “dose” of headings so to speak—to the changes that we detect in their brain structure and function when we examine them using DTI and cognitive tests.

Soccer players seem to be a good model for another reason. It’s becoming increasingly apparent that multiple head injuries lead to effects that go beyond what you’d expect from just adding individual events. There is mounting evidence for a cumulative or what I like to call a “super-additive” effect of multiple head impacts. And in soccer, heading is done with remarkable frequency. The amateur soccer players who were the subjects of our first study reported heading soccer balls up to 5,000 times in the prior year.

DF: What did you find in that first study of soccer players, which you published in 2013?

MLL: We found that more heading was associated with lower FA in the white matter, as well as worse performance on cognitive tests—particularly memory tests.

Note that these are not people with a clinical diagnosis of TBI. These are regular folks who play a lot of soccer but also go to school, go to work, and seem to function normally. So it’s pretty remarkable that in this population the amount of heading relates to areas where brain structure is affected, and even more so impairments in functioning.

So that was really the first part of that story—simply that those relationships exist. The next part was that these relationships didn’t appear to be simple linear relationships. In other words, people who reported fewer than about 1,000 headers in the previous year generally stayed in the normal range on brain imaging and function tests. But those who exceeded 1,000 headers per year had on average a dramatic reduction in FA as well as in cognitive performance. So this was at least a preliminary indication that there may be a threshold above which the risk for adverse effects on the brain goes up steeply—and that those who can keep below that level of exposure may be relatively safe.

Now, that’s not the end of the story, because we observed that a subset of individuals who did very little heading, looked, on our tests, like the people who did a lot of heading. The number of people in this study was too small for us to draw strong conclusions, but this observation at least points to the possibility that a minority of people are more sensitive to repetitive head injury, and may have significant adverse effects even with modest exposures.

DF: Could that extra-sensitivity explain some of the cases in the news—pro football players who decline cognitively and commit suicide, even college and high school players with no record of repeated concussions?

Yes, potentially. It needs to be determined in each case whether it’s a sensitivity or just a massive exposure to head impacts. But you can imagine using an approach like ours to define a risk factor—perhaps a gene—that confers sensitivity. Then you could advise people who have that factor not to engage in collision sports or go into combat, or otherwise expose themselves to head injuries.

DF: What are you doing to follow-up on that first study of soccer players?

MLL: We’re doing a larger study and tracking the players and their heading exposures and brain tests over time. Initially it was going to be a study of 250 players, each followed for one year, finishing in 2015, but we were fortunate to get substantial additional funding from NIH, so now we plan to finish in 2018, at which point we will have followed over 400 players for at least two years each. It’ll be a more substantial chunk of data from which we can draw inferences. One area we will begin to address is the role of genes that may predispose players to worse effects of heading. For example, the apolipoprotein E gene, in certain variants, is known for its association with degenerative diseases such as Alzheimer’s disease. In that light, we will explore its potential role in the evolution of brain pathology following mild trauma.

DF: Are researchers getting a better sense of how big a public health issue mTBI represents?

MLL: In terms of concussion, if you look at the data collected by the Centers for Disease Control and the World Health Organization, there’s a pretty good sense that there is a big problem. Even so, those data are likely to dramatically under-represent the problem. The large epidemiological studies are based on self-reporting or professional diagnosis of concussions, so the estimates of the incidence or prevalence of concussion or TBI will depend on the recognition of those injuries. If we don’t even know that they’re occurring, then they won’t make it into the statistics.

The people at the mild end of the TBI spectrum are much more likely to have their injuries go unrecognized. If a person isn’t knocked out cold, there’s a good chance even in 2015 that he won’t be seen as having had a concussion. There’s also a lot of variability out there in the way that athletic trainers, coaches, and physicians recognize a concussion. People don’t always understand that when someone has an impact to the head playing football, and “sees stars” and is a little confused, that’s actually a concussion—it may be a mild concussion but it’s still a brain injury.

And now we’re seeing evidence that multiple events like that, in quick succession, can have especially adverse effects. So the most important piece of advice for someone who has just been concussed is to avoid hits to the head again until he’s had time for a full recovery. But if he doesn’t recognize the first concussion, he won’t be careful to avoid the second one—he’ll go right back onto the field and run the risk of being hit again and again.

DF: What about treatments?

MLL: Although it is a massive and growing public health problem, TBI doesn’t have a specific treatment. Some potential treatments have looked very promising in preclinical studies in lab animals. But despite multiple attempts, most recently the ProTECT trial of progesterone, there has never been one that demonstrated efficacy in a clinical trial in people.

Why has it been so hard to find an effective treatment for TBI? I would say that one likely reason is that TBI is a very heterogeneous disorder. People are affected differently. And fortunately, most people will recover pretty well on their own. The people that we’re really worried about, the ones with bad outcomes, are a minority of the total group. So in studying prospective treatments for TBI, we need to be able to identify the people who most need the treatment.

In other words, if I have 1,000 people who come into my emergency room with concussion, I can be pretty sure that 700 or 800 of them at least will have a good outcome without treatment. It’s only the remainder who won’t do well without treatment. If I have a new treatment idea and I give that same treatment to all 1,000 people, I may be completely unable to demonstrate an effect of the treatment, because the majority of the people in the study are going to have good outcomes regardless of whether they are treated. So a big role for this research is to define the individuals who are most in need of treatments, so we can focus on this high-risk group in our clinical trials. I think that the type of study we’re now doing, with brain imaging, is going to be one way of making that distinction.