When he first saw the data, Patrick Bellgowan figured he had made a mistake. How else to explain it? A scientist at the University of Tulsa's Laureate Institute for Brain Research, Bellgowan had been scanning the brains of college football players, comparing the results to scans of non-players. His focus was on the hippocampus, a seahorse-shaped area deep inside the brain that plays an important role in emotional control and memory formation.

Bellgowan knew that the hippocampus was sensitive to traumatic brain injury. That shrinkage of the region was a hallmark of Alzheimer's disease. That a similar reduction in hippocampal volume also corresponded with chronic traumatic encephalopathy (CTE), a neurodegenerative disease linked to football, boxing and other contact sports. Still, he expected to see little difference in hippocampal size between a group of 50 college players and 25 non-players, mostly because everyone in the study was relatively healthy and young.

And then Rashimi Singh, a fellow researcher at the Laureate Institute, brought a set of initial results to Bellgowan's office.

"Are you sure?" he said. "That can't be right."

The numbers were stark. A group of 25 players with no history of reported concussions had hippocampuses that were, on average, 14 percent smaller than those of a control group of 25 males of similar age and health who didn't play contact sports.

Moreover, the same brain region in a second group of 25 players who had suffered at least one clinically-diagnosed concussion was, on average, 25 percent smaller than the control group -- a larger difference in volume, Bellgowan says, than the variations scientists have observed between the brains of healthy people and patients suffering from Alzheimer's or severe depression.

"I can't tell you how many times we checked this over," Bellgowan says. "The effect size was really large. It was really surprising."

Published this week in the Journal of the American Medical Association, the results of Bellgowan and Singh's research may be unexpected, but they also reflect a larger scientific trend. While much of the health and safety debate over football and other contact sports focuses on the risk of developing severe, headline-grabbing neurodegenerative diseases like amyotrophic lateral sclerosis (ALS) and CTE, a growing body of evidence suggests that both concussions and subconcussive blows can alter mood, cognition and behavior while causing damage and structural changes to the brain.

In other words, contact sports may be worse for your cognitive health than previously assumed, even if you don't ultimately end up in a dementia ward. Consider:

Of course, the above studies come with serious caveats: small sample sizes, a need for replication, a focus on correlation instead of causation. Much remains unknown. While scientists agree that getting hit in the head can result in both brain changes and damage, they don't always know if the former indicates the latter -- and even when it does, the extent of the harm remains unclear. Likewise, researchers are still sussing out the exact mechanisms by which changes and damage affect behavior and cognition, as well as the brain's ability to compensate for injury and/or heal itself.

Some towns are cutting high school and youth football due to worries over safety and brain injuries. (Getty Images)

Take Bellgowan's study. The authors acknowledge that absorbing repeated blows to the head while playing football may not be the cause of hippocampal differences. Genetic variations, environment and high levels of stress hormones may all play a role. Bellgowan suspects that football-induced head trauma is a key factor, mostly because the difference between players and non-players is so pronounced. He also says that the inverse relationship between reaction time and hippocampal volume and years played points toward a cumulative effect -- more hits producing greater damage over time.

If Bellgowan can secure funding for a follow-up study, he wants to study younger players, use more sensitive neuropsychological tests to screen for subtle cognitive impairment and test his hypothesis that the body's inflammatory markers may be crossing the blood-brain barrier and causing or influencing hippocampal changes. Looking forward, he'd like to study NFL retirees, too, the better to get a cross-sectional picture of how brain changes develop -- or don't develop -- over time.

In diseases like Parkinson's and Alzheimer's, hippocampal shrinkage can take place years before symptoms emerge. Do the differences observed by Bellgowan in college football players presage long-term disease? Or permanent structural harm?

Without more data, no one knows.

"The hippocampus can grow new cells," Bellgowan says. "These guys are young. So they might recover and come back once they stop playing if it's football-related. That's something we are going to try to answer. It's challenging one to pursue.

"In more severe types of TBI, you see hippocampus dysfunction and volume reduction. And with mild TBI, you've also seen that in older [football] players. What separates this study is that there are average 20-year-old males, still in the process of neurodevelopment. That is where we are driving. We think the key is to get at the developmental stage. This must be intersecting with neurodevelopment."

What does this uncertainty mean for contact sports? It means leagues such as the NFL have a choice. As do athletes and their families. And the rest of us. We can look at "League of Denial," stories of athletes whose lives have been ruined by brain damage, Bellgowan's study and others like it, a growing body of suggestive but inconclusive evidence, and conclude that, well, we just don't know enough. Not yet. So let's not panic or do anything hasty. Make a few adjustments, but keep on keepin' on. Alternately, we can evaluate the exact same set of incomplete facts and conclude that, well, we just don't know enough. Not yet. So let's slow down or stop until we know what's happening. Better safe than sorry.

So far, we've seen more of the former than the latter. Particularly from the NFL. On one hand, the league has warned its players about concussions, taken steps to better identify and remove concussed players from the field, publicly acknowledged -- one time, and only one time, but still -- the link between football-induced brain trauma and long-term harm. On the other, commissioner Roger Goodell continues to make noise about expanded playoffs and an 18-game regular season, both of which would expose players to more hits to the head. The NFL supports a proposed concussion lawsuit settlement containing brain damage evaluation and compensation provisions that do not account for future scientific advances nor offer relief to former players suffering from impairment that fails to rise to the level of neurodegenerative disease. It also endorses tackle football for children through a league-funded "Heads Up" program that purports to make the sport safer through altered tackling technique, never mind a lack of evidence showing that the program works.

During a London visit last year, Goodell told reporters that "the game isn't bad for you -- there are injuries and you need to recover from them, it's like anything" -- a dubious proposition -- while NFL Head, Neck and Spine committee co-chair Richard Ellenbogen recently told an International Olympic Committee injury prevention conference that "media is not the place to debate the science on concussions." (An even more dubious proposition, given that public debate is what pressured the league to do anything beyond denying and dissembling about concussions in the first place, and that Ellenbogen likely wouldn't have his current gig without media exposes of his predecessors, Ira Casson and Elliot Pellman).

The other week, Bellgowan says, one of his son's lacrosse teammates suffered an obvious concussion. The boy had trouble balancing. Could barely walk. Concerned, Bellgowan approached the boy's father.

You really need to see a doctor, Bellgowan said. Gotta hold him out. Go see a specialist.

Yeah, yeah, yeah, the father replied dismissively.

"I can't tell you that head hits caused the [hippocampal volume differences] seen in our study," Bellgowan says. "But certainly, the correlation suggests that it may be one of the reasons. And because these guys are young and in the neurodevelopmental stage, instead of being mature adults, you should treat things conservatively. I always tell parents, 'if you miss a game but save a brain injury, that's good.'"

Is society willing to make that trade? To err on the side of caution in a way that protects people more than the games they play? Without more data, no one knows. But as disturbing research keeps piling up -- as the hits keep accumulating -- change seems inevitable. Shrinkage, too. Just this week, the New York Times reported that an East Texas town scuttled its seventh-grade tackle football program in favor of flag. Parents were worried about the safety of their children, and as a town lawyer told the newspaper, "there's too much evidence now." The next time a study indicates that contact sports may be worse for the brain than previously believed, nobody should be surprised.