How Does Playing Football Affect Your Health?

JENA: So, you know, I— I like really wild and crazy ideas. And a few years ago, you approached me with a— a wild and crazy idea.

VENKATARAMANI: This was on one of our walks, I think, Bapu, where we talked about this.

That’s my friend Dr. Atheen Venkataramani.

VENKATARAMANI: I’m a physician and a health economist at the University of Pennsylvania.

Atheen and I met in residency at Mass General Hospital, and now we have similar jobs. Unlike me, though, Atheen is a big football fan — he’s been rooting for the Pittsburgh Steelers since he was a kid. And like a lot of football fans, he’s been increasingly aware of the toll the game takes on the bodies of the people who play it.

VENKATARAMANI: Most of the discussion we’ve had up until the last few years has been with the neurological health of football players. More recently we’re thinking about, the stress on the heart that may be adaptive for the game, but it may or may not be adaptive for life after that.

Are years of stress on the heart, repeated blows to the head, and other physical injuries causing N.F.L. players to die earlier than they otherwise would have? What about the health advantages that come with being a professional athlete?

So … can we just look at health data for N.F.L. players and compare them to the rest of us? The problem is that they aren’t like the rest of us.

VENKATARAMANI: On average, football players live a lot longer than we do. There’s a lifetime of exercise and preparation and discipline that can have health benefits. And so you wouldn’t want to necessarily try and understand whether professional football can be harmful by comparing a professional football player to you or I or any regular Joe that you know. You need a better comparison.

A few years ago, that’s what Atheen and I set out to find. I’m Bapu Jena, and this is Freakonomics, M.D. Today on the show: How does pro football really affect your health?

VENKATARAMANI: Back in 1987, the N.F.L.’s Players Association held a strike because they were unhappy with some of the rules around free agency and they left for 24 days, about three football weeks. And the N.F.L. owners decided to continue playing N.F.L. games, but they didn’t have players. So they had to find players. And basically local teams went to advertise in supermarkets. They held tryouts. They looked for ex-college players that lived in the area. And they formed these teams of what are called replacement players. it’s kind of a famous episode in N.F.L. history. It was made into a movie called The Replacements.

CLIP: Here’s a list of people I’ve been keeping my eye on over the years. They’ve all played football somewhere, not all of them in the pros. But they all have something unique to bring to the game.

VENKATARAMANI: The replacement player situation gave us this really interesting natural experiment where you had people that wouldn’t have been in the N.F.L. otherwise play a couple games. And as a result of that, they were recorded in official databases. And so now you have a comparison group of people that credibly were really, really good at football, just didn’t play professionally. And you could compare them against people that actually did play in the N.F.L.

JENA: You could imagine a totally different approach where you look at N.F.L. players and study the relationship between how long they played in the N.F.L. and their health outcomes. So if someone played three years or five or 10, maybe you compare these players and draw conclusions about football’s effect on their health based on what you find. What’s the problem with that strategy though?

VENKATARAMANI: The problem there is people that tend to play longer, we know, tend to actually have a survival advantage because whatever it is that allows them to play longer also allows them to live longer. And so this is something that we can call the healthy worker effect. The people that are working longer tend to be healthier. It’s well described in epidemiology. And you see it in industries like the N.F.L., but you can also see it in occupations like coal mining.

JENA: So if you take N.F.L. players and you compare them to replacement players, what do you find?

VENKATARAMANI: The study we did, Bapu, we had about 3,000 N.F.L. players and almost 900 replacement players. We basically looked at cohorts of players that entered somewhere for their first game between 1982 and 1992. We followed them up to the end of 2016. And you do see that N.F.L. players tend to have shorter survival compared to replacement players because, you know, these people are still relatively young. There aren’t a lot of deaths in the sample, so we can’t be a hundred percent sure that that difference will hold up if we were to, say, redo the study today or five or 10 years from now. But at the time we did it, we saw the signal that N.F.L. players had a 38 percent higher risk of death compared to a replacement player as you follow them through 2016.

A 38 percent higher risk sounds high — but we can’t be sure that it’s meaningful. Even though our sample size was fairly large, at around 4,000 people, only 144 players and 37 replacement players died before the end of 2016. That difference isn’t quite statistically significant.

VENKATARAMANI: We just didn’t have a lot of events, in this case, deaths. And if you don’t have a lot of deaths, then predicting what causes those deaths just becomes harder, and it’s a noisier exercise.

JENA: So there was this mortality signal that was I think quite interesting. What else was there?

VENKATARAMANI: The second piece is we could generate a confidence interval for the relative difference in mortality between N.F.L. players and replacement players

A confidence interval helps researchers convey just how precise a measurement is, or isn’t. It helps us rule out findings that are too large or too small. In this study Atheen and I are talking about, the confidence interval revealed something interesting.

VENKATARAMANI: It entirely rules out the survival advantage that is seen in other studies that compares N.F.L. players to the general population. So we can say with confidence that that survival advantage is not actually something that materializes when you have a more fair comparison.

So, just to be clear: the survival advantage for N.F.L players that has been found in other studies is actually sort of an illusion. If you compare professional football players to average people, they seem healthier. But that’s in part because N.F.L. players are often in terrific physical shape. They also have access to top-notch health care, and make a lot of money. All three of those factors are associated with better health outcomes. What Atheen and I saw in our study was that when you compare N.F.L. players to a much more similar group — like the replacement players — they didn’t live any longer. But there was more.

VENKATARAMANI: We took this incredible crowdsource data source called Pro-Football-Reference.com. It’s a database that I am completely obsessed with as a football fan. That data set has information on where people are born, their full names, and their exact date of birth. So you can use those pieces of information to search the National Death Index. We ran it through 3,800 people looking to see if we could find a matching death certificate That took some time, but I’m glad we did it, Bapu, because we were able to find some pretty interesting patterns. What we found was that seven of the N.F.L. players in our sample died of A.L.S. And there were no such deaths like that in the replacement player sample. If those two numbers were statistically similar, we would expect to seeat least two or three such deaths in the replacement player sample, and we didn’t.

This study that Atheen and I worked on, along with Dr. Maheer Gandhavadi, was published in the Journal of the American Medical Association in 2018. It was among the first to suggest that playing in the N.F.L. does not confer a health or mortality advantage — despite the physical fitness the sport demands and the lifestyle and health care it affords participants. We’d probably learn a lot more if we re-did that research now, or in another 10 or 15 years.

But, how else could we figure out if playing football is good or bad for people’s health? After the break, you’ll hear us brainstorm some good, and bad, ideas. I’m Bapu Jena, and this is Freakonomics, M.D.

*      *      *

JENA: So Atheen, I— I’ve got some ideas. I want to get your input on what you think about them.

VENKATARAMANI: Yeah, let’s do it.

One of the best parts of my job is coming up with ideas for studies, creative ways to solve a problem. In hindsight the idea we land on usually feels obvious, but getting there is hard. When I sit around with my colleagues and students and hunt for our next idea, it sounds a little something like this:

JENA: One factor that I’m always interested in that’s outside of our control is weather. You could imagine some players might be more likely to suffer an injury early in their career if they’re exposed to bad weather conditions. And as a result, they were injured and had shorter tenures in the N.F.L. Could that be a natural experiment, to look at how exposure to bad weather early in a player’s career impacted their injury risk, and therefore how long they played? Because that gives us a totally random reason for the injury that’s unconnected to a player’s underlying health.

VENKATARAMANI: You would need a lot of events. And we’d need to show basically that when these events happen in critical times in people’s career, like their rookie year, that that does shorten tenure. So I think it’s super creative. My skepticism is that I’m not sure how strong that signal will be in predicting the length of tenure.

JENA: The other idea I had is the following: are there scenarios where there are plausibly random or exogenous things that affect injury rates? I don’t know anything about football, by the way, but if you happen to be a player against a team where in that year the defensive players are more likely to cause injury, could we use that variation to say, all right, there are some instances where players are more likely to be injured than otherwise similar players who by chance weren’t exposed to defensive players more prone to cause injury? Does that make sense, what I’m saying?

VENKATARAMANI: Yeah, I think it does. It reminds me of this Michael Lewis book, The Blind Side. It’s about this guy Michael Oher, but it talks about how the person protecting the quarterback’s blind side had to become bigger and more athletic over time. And one of the things that I think really sped up that evolution was the presence of Lawrence Taylor on the New York Giants, who was a completely different force than defensive players had been in the past. And the game had to respond to him, and it did over time. But in that intervening period, it’s possible that the adjustment that didn’t happen led to some injuries that may not have happened.

It also reminds me of the old Veterans Stadium before it became the Linc where the Eagles play was kind of a disaster to play on. They had parts of the field that were raised. It was this like weird AstroTurf. So it was this place where injuries were kind of waiting to happen. And the schedule is not set by the players. It’s set by the league. So to the extent that you’re exposed to environments that may or may not be relatively less safe, that’s something you could exploit there too. Bapu, here’s another thing There are a lot of N.F.L. players that were good at another sport in high school or college but at some point in their lives, they chose to play football. So there is this possibility of comparing football players to professional athletes in other sports. And a paper that I like a lot and kind of follows the spirit of what we were trying to do, is one that actually did this comparison of N.F.L. players to baseball players. They had bigger sample sizes than we did. And in their comparison, they found that N.F.L. players have shorter life expectancies than M.LB. players.

JENA: You know we’re talking about the N.F.L. And it’s of interest to a lot of people, but I think about the decision that parents may make about whether or not they want their child to play high school or college football. That gives me a couple other ideas. You can imagine if there are communities where a child is hurt severely in a football game that event would reduce enrollment in football leagues in that area compared to otherwise similar areas where, such an event didn’t occur. And could you then look at outcomes of kids who might have had a high probability of playing in football? So that would be one approach.

VENKATARAMANI: What do you think of this, Bapu? So, running a high school football program costs a lot of money and sometimes you have to drop your football program. So there are high school students that, you know, by virtue of when they were born, may or may not have had access to a football team. Do you think that’d be a way to kind of find a fair comparison for participating in high school football versus not?

JENA: I think that would be a terrific approach. I think the challenge there would be how do you identify the people who are the relevant group to study? Because you can’t say, let’s look at all 500 boys in the pre period in that school and 500 boys in the post period because only say 10 or 15 of them would’ve played football. But I like that idea. All right, so I have one last question for you, Atheen. Knowing what you know, would you let your child play football?

VENKATARAMANI: It’s so hard for me to answer this question because I love football so much It’s such a lovely game. There’s so much strategy. There’s so much heart. And part of the reason that we did the paper, Bapu, was I just wanted players to be informed about the risks. But yeah, if I were to pick for my son, that would be right now a no.

JENA: I probably would say the same thing. But it is really important not to understate the other benefits that come from playing in a team, like the camaraderie, the self-confidence, the exercise, the discipline. As an economist, I’d be thinking, well, can you have your cake and eat it too? You could get all that stuff from playing on a tennis team or a soccer team or a baseball team. And this is an area where we probably lack information for a decision that many, many, many people every year are making. We know the pros. But we don’t really have a good insight into the negatives.

VENKATARAMANI: I 100 percent agree.

That’s it for today’s show. I’d like to thank my guest, Dr. Atheen Venkataramani, for his time and his honesty. And thanks to you, of course, for listening. If you can’t tell, I like to think out loud and we’re gonna try to do a lot more of that moving forward. Let me know what you think about that. I’m at bapu@freakonomics.com, that’s B-A-P-U at freakonomics dot com.

Coming up next week on the show: I’ll explain how a government program that ended 70 years ago can help us answer some really important questions today — about what we eat, and about a lot more. If you’ve got a sweet tooth, you won’t want to miss this episode! That’s next week on Freakonomics, M.D.

*      *      *

Freakonomics, M.D. is part of the Freakonomics Radio Network, which also includes Freakonomics Radio, No Stupid Questions, and People I (Mostly) Admire. All our shows are produced by Stitcher and Renbud Radio. You can find us on Twitter at @drbapupod. This episode was produced by Julie Kanfer and mixed by Eleanor Osborne. Lyric Bowditch is our production associate. Our executive team is Neal Carruth, Gabriel Roth, and Stephen Dubner. Original music composed by Luis Guerra. If you like this show or any other show in the Freakonomics Radio Network please recommend it to your family and friends. That’s the best way to support the podcasts you love. As always, thanks for listening.

VENKATARAMANI: My son is also built like me, so, , I don’t know. That’s not—.

JENA: Like a linebacker? Or— or like a data programmer. Which one?

VENKATARAMANI: The second thing.