At the youth level, a child who played American football in 2013 had a 3 percent chance of sustaining a concussion during a single season. That number was a little higher at the high school and college football level, however, where players had a 5 percent and 6 percent chance of sustaining a concussion, respectively. These risk estimates come from the first study to compare the incidence of concussions across the entire range of youth football — that is, for kids between the ages of five and 23.

"The risk of concussion is lowest at the youth level, and increases with each successive level of competition," says Thomas Dompier, an epidemiologist at the Datalys Center for Sports Injury and Prevention and a co-author of the study, published in the medical journal JAMA Pediatrics today.

In the study, researchers analyzed injury data collected by athletic trainers at youth, high school, and college football levels during both games and practices in 2012 and 2013. They found that concussions accounted for 9.6 percent, 4 percent, and 8 percent of all youth, high school, and college football injuries, respectively. Overall, games posed the highest concussion risk at all levels, but the rate of concussion during games was significantly higher at the college level. Interestingly, the lowest concussion rate during football practice was also found at the college level.

These results suggest that "more can be done at the high school and youth levels to mitigate concussion risk," Dompier says. Educating coaches, implementing policies aimed at reducing the amount of full contact drills, and increasing player knowledge of proper tackling technique could help reduce these numbers, he says.

The study, which was partly funded by USA Football — the governing body for amateur football — and the NCAA, used data collected from thousands of players across the US. Still, it's possible that it underestimates the number of concussions sustained during 2012 and 2013, especially those at the youth level. One potential issue lies with the risk calculation; it was calculated by dividing the number of players who sustained at least one concussion by the number of players who started the season. But kids drop out of sports all the time — especially at a young age. This means that the overall risk may be higher than the study indicates.

In addition, all injuries were recorded by athletic trainers, which means that the injury reports were largely dependent on a trainer’s experience. Coaches who are less experienced may not be as good as detecting the signs of a concussion compared with coaches who have been working with youth for a long time, the authors note. This is most apparent in the five to seven age range, where no concussions were reported. Sure, it’s possible that these kids actually sustained zero concussions during two seasons. But it’s also possible that very young children aren’t taught to recognize and report their symptoms. Trainers may also be less skilled at recognizing the symptoms of a concussion in young kids.

Still, athletic trainers appear to be getting better at recognizing concussions. The rate of concussions in youth football players recorded in this study was higher than the rate that Dompier recorded in a comparable 2007 study. This suggests that coaches at that level are getting better at detecting and reporting concussions. "I believe this may be reflective of educational efforts to inform players, coaches, and parents on the signs and symptoms of concussion," he says. The rates observed at the high school and collegiate level were comparable to previous studies, however.

Dompier and his team aren't just interested in football; they are currently replicating this study in boys and girls youth lacrosse. They also just completed a study that looked at the effectiveness of policy changes and coach education in reducing injury risk and head impact exposure in youth football; it will be published soon, Dompier says.

Despite these efforts, there's a lot more work to be done in this area, notably in female sports, which "have been underrepresented in the literature," Dompier says. The researchers hope to make similar comparisons across the entire spectrum of youth in other boys and girls sports, such as soccer, ice hockey, and basketball.

Verge Video: Can a concussion sensor help save athletes' lives?