YOU WOULDN'T GUESS that Jefferson High football players Joel Ripke and Brandon Stumph are part of a scientific breakthrough. Purdue researchers who put sensors in the helmets of the seniors from Lafayette, Ind., certainly didn't. Ripke, a mountainous 17-year-old at 6'6" and 260 pounds, is the Bronchos' starting right tackle. His buddy Stumph, a starter at defensive end, is a more mundane 6'1" and 190 pounds, but with a thirst for contact. His black helmet looks like one of those chipped and gouged bowling balls that hasn't beaten a straight path in years.
Despite their easy camaraderie and Penn-and-Teller size difference, Ripke and Stumph line up across from each other in practice and get after it, with Stumph breaking out every duck, dodge, chop or bull rush he knows to get past Ripke's forklift arms and Frisbee-sized mitts. "If I'm not bigger than the dude, I like to hit him with my helmet," Stumph says, "and then use a move so I can get his hands off me." Nor does Ripke shy away from putting hat on hat. He's been taught that effective run blocking requires three points of contact on the defender: hand, hand, helmet.
Despite their frequent bell-ringings and clock-cleanings, neither Ripke nor Stumph has suffered a concussion in practice or in a game. That would be unequivocally gratifying news, except that the Purdue researchers' data, to be published in the Journal of Neurotrauma, tell a far more troubling story. The findings suggest that while the NFL is going to unprecedented lengths to control the violent collisions that produce concussions, brain trauma in football may start much earlier, and much less conspicuously, with hits that never raise an eyebrow, much less a penalty flag.
Before the 2009 football season the group of Purdue engineering professors, athletic trainers and graduate students fitted 23 of the Bronchos' helmets with accelerometers and gave players both the ImPACT test—a computerized neurocognitive exam that tests memory and concentration—and tests of working memory while their brains were monitored with magnetic resonance imaging (MRI). The idea was to establish a baseline for each player against which he could be reexamined after a concussion. Says Thomas Talavage, a Purdue associate professor of biomedical engineering and electrical and computer engineering, "We were looking to understand what kinds of hits cause a concussion and what the consequences are."
Using NFL-sponsored studies as a guide, the researchers figured that hits in excess of 80 times the force of gravity (heading a soccer ball produces around 20 Gs) would cause concussions. So the Purdue researchers were stunned when, on the first day of full-contact practice, they started seeing hits of 100 Gs or more. "I thought, Oh, my god, we're going to be carrying these kids off the field," says Eric Nauman, associate professor of mechanical and biomedical engineering.
It turned out, however, that no particular magnitude of hit correlated with a concussion. One player holding the line on an extra-point attempt took 289 Gs to the helmet from a converging pair of would-be kick blockers. "You could hear the hit in the subdivision next door," says Evan Breedlove, a biomedical engineering grad student and member of the study team. But the lineman was fine. In fact, three weeks into the season the Purdue team had just one concussion for its study. (There were concussions among Bronchos players who were not part of the test.) So the researchers had players from the study who had never suffered concussions retake the ImPACT test and get their brains scanned with functional MRIs (fMRI), which image cerebral blood flow to pinpoint active areas in the brain. The players were meant to serve as a control group for later comparison to concussed teammates. But the first lineman who came in as an ostensible control subject surprised the researchers when, compared with the preseason, he scored 20% lower on the visual memory section of the ImPACT test, which requires rapid identification of recurring patterns. The player had no trouble with the verbal section, though, and Talavage began to think there might be something wrong with the test itself, which is used by the NFL and many college and high school teams to gauge whether a player has recovered from a concussion.
A few concussions did arise as the season went on, but the researchers continued to bring in nonconcussed players for ImPACT tests and fMRIs. And then they saw it again: Another kid who had never suffered a concussion flubbed the visual memory section of the ImPACT test. Of 11 players who took midseason testing, three had suffered concussions during the season and eight had never had concussions. Of those eight, four nevertheless showed significant declines in visual memory. In fact, the players with the most impaired visual memory skills were not coming from the concussed group but from a group that in the week preceding the test had taken a large numbers of hits—around 150—mostly in the 40 to 80 G range.
If the test scores were accurate, the researchers had inadvertently documented, in real time, a new classification of high school athlete: a player who was never concussed, was not verbally impaired and was asymptomatic even as far as his parents could tell, but whose visual memory was more impaired than his amnesic, headachy, light-sensitive, concussed teammates.
Says Talavage, "We started having weekly meetings to debate whether we were seeing something real."
AND THEN THEY looked at the fMRIs. Those brain snapshots had been done while players took two versions of a working memory test. In the first version a subject must click a button each time a flashing letter repeats in sequence. D, A, B, B—click. The second version requires more brainpower: React when the letter that flashed two characters ago repeats. A, J, F, J—click.