Super Bad: Does an NFL Career Guarantee a Brain Disease?

Another study reports an alarmingly high rate of CTE among NFL players, but, once again, it overstates the case.

Just in time for the Super Bowl, we NFL fans were treated to another reminder about our favorite sport: it’s bad for one’s health. To be more specific, the same folks at Boston University who have been publishing research on Chronic Traumatic Encephalopathy (“CTE”) for a decade have autopsied more brains, and found that 92% of the 376 former NFL players studied had evidence of CTE, the degenerative brain disease that can strike former players both early and late after their playing careers, often with dramatic behavioral changes and profound memory issues. Sounds alarming, right? The problem is that the framing is neither reasonable nor helpful.

Of course, news media ran with the headline:

Fox Sports, I suppose, can be forgiven for seeking clicks with disturbing NFL news in the week before the Super Bowl. However, Boston University did not do better with its own press release:

(Speaking of trauma, apologies to any still-grieving Eagles fans for the photo.)

Rather surprisingly, none of the major media outlets that I found running the story emphasized the obvious flaw of the study: we have a numerator in search of a denominator. Yes, 345, or 91.7%, of the 376 NFL former players’ brains autopsied showed evidence of CTA; however, were the 376 players selected at random, or somehow carefully curated from a broad sample to be at an average risk of CTE? Of course not!

These were players for whom family members donated their brains to what is now known as the UNITE Brain Bank, whose mission as clearly stated on their web site is to study brains affected by repetitive trauma:

You might conclude that the average family who sends their beloved NFL player’s brain to be studied by this organization has one major concern: that their family member had CTE.

My point is not to trivialize CTE, or stanch the nausea I feel when, as a physician, I see an NFL player laid out by a hit to the head. However, people need to make their own choices around sports, and putting poorly-contextualized numbers into the world does not help those decisions.

When Dr McKee and her group released their prior major report in 2017 (finding CTE in 99% of former NFL players studied), many outlets were quick to point out the limitations in their study. Even Dr McKee admits the obvious now, but her somewhat understated words did not make it into the BU presser, but rather an interview with CBS:

Alarmist headlines aside, then, how common really is CTE among football players, and is there anything that can be done to reduce its incidence? Many researchers, including Dr McKee, have been grappling with these questions, without much success.

Even FiveThirtyEight took a stab at generalizing CTE rates back when the 2017 BU research was released, finding that 99% rate of CTE among former NFL players’ brains. Borrowing from a Neurology paper, they noted the authors attempted to generalize from the entire pool of 1,142 NFL players who had died during the 8 year period while the brains autopsied were collected. If only half of players who had died with CTE had been captured via autopsy, that would mean almost 20% of all the players dying in that time frame would have had CTE on biopsy; if a higher proportion like 90% were donated to the brain bank, that number would drop to around 10% (the authors suspected the true rate was in the 20-30% range). Also noted was the concern expressed by other researchers that perhaps the bar for a CTE diagnosis via biopsy has been set too low, and Dr McKenna’s team is over-diagnosing CTE. We lack agreement as to just how much of the mis-formed tau proteins associated with CTE are adequate to make a diagnosis.

Sometimes, that most unscientific of techniques, the “eye test,” can nonetheless be helpful when numerators are being discussed in the absence of reliable denominators. One such example (in this case a numerator issue): the studies claiming Long COVID to be in the 30+% range of all Covid cases are patently ridiculous to any physician, like me, with close follow-up of hundreds of Covid infections and a mere handful of Long COVID illnesses. When a study with terrible limitations creates a data point that is off from professional experience by an order of magnitude or more, it’s probably fair to focus on the study’s limitations rather than its conclusions.

Here, though, it’s trickier. Who has a large cohort of NFL players to follow? I have cared for exactly one retired NFL player (and, of course, he had been diagnosed with CTE). The NFL has this access, but being involved in eternal litigation on this subject, tends to be tight-lipped on the subject.

Clearly, these 92%, or 99%, figures are not representative of NFL players at large; we have too many out in the world functioning at a high level to even consider this as a true high end number. When I think about a discrete cohort with which I am quite familiar, the Hall of Fame core of the Steelers dynasty I cheered for as a kid comes right to mind. Of their ten players in the Hall of Fame from that era, only one (deceased center Mike Webster) has been diagnosed with CTE, and most of those players are moving through their 70s now in visible public roles without reports of signs of CTE, none more so than Terry Bradshaw, who was victim of one of the most unwatchable head injuries in the history of the NFL:

(I don’t really need to say anything about Cleveland Browns fans that they would consider this photo “merchandise.”)

So, 1 out of 10 in that very crude sample; while lacking any real statistical value, I will note that a cohort consisting of Hall of Fame players inherently is comprised of players with long careers and therefore more risk of head injuries. There is no way a number north of 90% is in play here.

Of course, the ideal cohort to follow is every single NFL player. Interestingly enough, in 2008 the NFL commissioned a study with researchers from the University of Michigan, who attempted to survey a “stratified random sample” of about 1000 of the 7000 living, retired NFL players eligible for pension. It’s worth noting, of course, that restricting the sample to living players could under-estimate the toll of CTE, given its propensity to shorten lifespan. As with any survey-based study, there is always the concern that respondents are not typical of the entire sample; here, about 65% of those contacted completed the telephone interview, and it is hard to know in which direction that might bias the sample, in terms of whether people affected by CTE might be more or less motivated and/or able to participate in the study.

In any case, the researchers did find a substantially higher rate of any sort of dementia among both younger and older former players compared to the general population, but rates were rather low overall (1.9% in those 30-49 years old, and 6.1% among those over 50). Depression was slightly more prevalent compared to the general population, but “Intermittent Explosive Disorder”, one of the concerns for early-onset CTE, was actually less common among former players than the general public. These are not exactly headline-grabbing figures.

CTE remains a risk that everyone would like to mitigate. Still unclear is how best to do so.

Do we need to change the way certain position groups play? The storyline exists that those in the trenches are at highest risk, due to the near-constant “subconcussive” helmet rattling between offensive and defensive linemen. We lack statistical evidence for this claim, however. The only study I could find even addressing the role of position in developing CTE hailed from Canada and was extremely small, with autopsies from only 25 former football players. Again, we see denominator issues, as all candidates were known to have had at least mild Traumatic Brain Injuries before their deaths. However, no positional patterns emerge from their findings; it’s also important to remember that almost half a roster is filled with offensive and defensive lineman:

Back to the eye test, any list of prominent cases of CTE victims tends to be a scattershot of positions; some of the most famous (running back Frank Gifford, tight end Aaron Hernandez, quarterback Ken Stabler, safety Andre Waters, linebacker Junior Seau) can be found all over the roster.

It’s unlikely, then, that rule tweaks to protect wide receivers, or change the way lineman interact, will rid football of CTE.

Helmet redesign is a popular deus ex machina for the problem. What a neat solution to simply change player helmets and make CTE go away. The problem, of course, is that nothing in life is so simple. I have long thought that making the NFL helmet less of a hardened weapon and more like a one-hit-and-break bike helmet would reduce players’ tendencies to lead with their helmet, and this concept has been explored via helmets with outside padding which, in a lab setting, seem to both reduce force of impact and better protect the wearer. One issue, of course, is that brain injuries are mostly caused by movement of the brain within the skull, and no helmet can greatly reduce these forces in a contact sport.

The terrible sequence of Miami Dolphin quarterback Tua Tagovailoa over weeks 3 and 4 this past season illustrates this concern; week 3 he was (rather gently) slung down to the turf, and the whiplash effect of his fall to the ground on his back would have been an issue even if his head had been packed in 6” of bubble wrap. The following week he was much more forcibly thrown to the ground, and immediately showed signs of what appeared to be a substantial brain injury from a similar mechanism, again almost impossible to prevent with a helmet.

An easy intervention like the use of the “Guardian” soft helmet cap seems reasonable, but still has its nay-sayers (mostly over the effect of its extra weight) despite some promising, albeit unscientific, early results on concussion reduction. These are unlikely to be a panacea, however, especially as tackling form continues to be dismal in the NFL — ask yourself how many times a game you see someone make a proper form “heads up” tackle with an arm wrap. They are more rare than my teaser bets hitting, I promise you.

Improving tackling form, penalizing hits to the head, limiting contact in practice, reducing kick returns, and, yes, tweaking helmet design, can probably all play a modest role in reducing brain trauma, and, ultimately, CTE; but nothing is likely to make it largely go away.

A question I am asked with some frequency, strongly related to this question of concussions and CTE, is: “should I let my kids play football?” We have little research on the subject of the CTE risk from playing only youth or high school football. The same Boston University group published a study from its 2017 autopsy data — so, again, denominator issues are in play — finding a strong correlation between years played and risk of CTE. “Only” 3 of the 14 athletes who had not played beyond high school football were found to have CTE on autopsy (all of them mild), a far lower rate than the 110 of 111 NFL players.

Of course, any increased risk of injury to an organ as precious as the brain is enough to give a parent pause. According to earlier research from Dr McKee, none of the 198 brains studied from people not involved in contact sports showed evidence of CTE, even though 33 of those subjects had suffered major brain injuries from assaults, car accidents, or other injuries. This implies that CTE is truly an avoidable disease, for those willing to avert long careers in sports like football, boxing, or hockey.

However, is a severe brain disease likely from a high school football career? A clever retrospective study from 2017 published in JAMA tried to answer that question in survey form, with all graduating high school football players from 1957 in the state of Wisconsin as the study cohort. Do bear in mind: this is the smelling salts/”What day is it? Right - get back out there, kid!” era of concussion mentality. Regardless, they found no difference at all in cognition at age 65, and significantly lower rates of depression among the football players. So, no — I don’t think four years of high school football will damage a teen’s brain; and, no, parents: despite your child being the strongest, fastest, and most talented kid in their class, that Division I offer and NFL career are just not going to materialize.

What about before high school? This one is murkier. The theory, proposed both by Dr McKee’s group and others, is that repeat trauma to still-developing brains increases risks of brain injury. The data is based on small studies, a mix of imaging research and behavioral questionnaires, and is not entirely compelling. That said — does a ten year old need to be playing tackle football? Flag football is probably just fine to scratch that particular athletic itch at a young age, and it tells me something that the NFL is fully on-board with the push for youth flag football.

Perhaps the question which I feel least able to answer is: “Why do only some people develop CTE?” Clearly there is an individual, if not fully genetic, component; otherwise the biggest hits or the most known concussions would consistently lead to the worst CTE cases. This is hardly true (witness my beloved Terry Bradshaw — and any comments questioning his cognitive ability based on his pre-game shows will not be entertained). A genetic predictor of CTE risk would be welcomed, but thus far has proven elusive. The gene heavily linked to dementia risk, APOEε4, has shown a questionable association with CTE; I suppose I might want to test my daughters for this gene prior to green-lighting joining the boxing team, but truly it is probably at most a small risk factor.

I think parents, and athletes, probably have to learn from their real world experiences. When I watch a player absolutely get their bell rung and pop right back up into the huddle, then see Tua Tagovailoa bounce off the turf and stagger like a punch-drunk fighter in the tenth round, I have to think: Tua probably is not meant for contact sports. My instinct is that athletes for whom small triggers lead to large concussions, or who suffer one or two severe brain injuries with prolonged symptoms (especially if not following particularly impactful hits), are letting us know their brains are vulnerable. I don’t know if that correlates to CTE risk, or any risk of long-term brain injury, but that might be the best we have in terms of “screening.”

To bring this back around, were I 30 years younger and had the physical and mental gifts of a Jalen Hurts, would I choose to play in the NFL, especially his style of football, knowing that I might have a 5 or 10%, maybe even 30+%, risk of causing myself a life-altering brain injury? Framed that way, it’s hard to answer in the affirmative.

But what if I just went to med school instead? A recent study found that 10% of the general population over 65 suffered from dementia, and over 20% from cognitive impairment — and that was just a slice of time among senior citizens, not a lifetime risk; dementia rates were 35% for people in their 90s. Hmmmm.

Spun that way, I’m reminded that no path in life is without its risks. I’d probably play. But I’d play for a proper team.