• John E Zvijac, Todd A Toriscelli, Shannon Merrick, and Gary M Kiebzak.
• UHZ Sports Medicine Institute, Doctors Hospital, Baptist Health South Florida, 1150 Campo Sano Ave, Suite 200, Coral Gables, FL 33146, USA.
• Am J Sports Med. 2013 Jul 1;41(7):1511-8.

BackgroundThere are conflicting reports regarding the association between isokinetic concentric quadriceps and hamstring strength deficits and ratios and risk for hamstring injuries in athletes.PurposeTo determine if isokinetic concentric Cybex data collected during the annual National Football League (NFL) Scouting Combine are predictive of hamstring injury in professional American football players during their first season.Study DesignCase-control study; Level of evidence, 3.MethodsAll 32 NFL teams identified players selected during the first 5 rounds of the NFL annual draft who had hamstring injuries during their first professional season. Of these, 164 players with 172 injuries also had Cybex data from the previous year's Combine. Analyses compared injured legs with contralateral uninjured legs and also injured players with uninjured controls using a database of Cybex data from all players who participated in the NFL Scouting Combine from 2006 to 2011.ResultsNo Cybex strength variable differentiated the injured legs from the contralateral uninjured legs or injured players from uninjured controls, even after taking into account days lost from activity. Mean ± SD peak torque for the injured and contralateral uninjured sides was as follows: 315.7 ± 70.0 and 313.5 ± 68.3 N · m, respectively (P = .773, paired t test), for quadriceps and 203.0 ± 42.4 and 205.3 ± 42.5 N · m, respectively (P = .608, paired t test), for hamstrings. The sensitivity and specificity for the hamstrings-to-quadriceps ratio predicting hamstring injury were 0.513 (95% confidence interval, 0.419-0.607) and 0.524 (0.495-0.554), indicating that the hamstrings-to-quadriceps ratio was not a useful predictor of injury (calculation used the mean ± SD ratio for injured legs, 0.656 ± 0.133). Side-to-side peak torque differences were also not predictive of injury, with more than a 10% difference (plus or minus) occurring commonly in both injured and uninjured players for quadriceps (53% prevalence for both injured and uninjured) and hamstrings (injured, 41% prevalence; uninjured, 43% prevalence).ConclusionIsokinetic strength data collected from collegiate players at the NFL Scouting Combine were not useful for predicting risk of hamstring injury in subsequent professional NFL competition. These data call into question commonly accepted assumptions about risk factors for hamstring injury, at least for elite collegiate American football players, when the temporal relationship between Cybex testing and actual injury is several months.

18 keywords...

hide…