Assessing the biomechanical risk factors for hamstring injury during maximal velocity sprinting

Introduction: In maximum speed running, the muscle load of hamstrings is greatest during the terminal swing phase of the sprinting cycle, when most hamstring strain injuries (HSI) occur. The assessment of biceps femoris long head (BFLH) muscle architecture, which is the most injured muscle, is usually conducted using ultrasound (US) with extrapolation methods often required to estimate the fascicle length (FL). Extensive literature also exists on Nordic hamstring exercise (NHE), a dynamic hamstring eccentric exercise, which has been proved to be effective both as a prevention and rehabilitation method for HSI. Research studies have examined various risk factors in sprinting, but little evidence exists on the relationships between these factors and the differences between males and females.

Aim: The primary aim of this research was to identify relationships between hamstring injury risk factors and to investigate whether these relationships differ between males and females. A secondary aim was to evaluate the agreement between extrapolation and direct measurement of muscle architecture when using US and, to quantify the acute adaptations of BFLH after conducting the NHE. 

Methods: Seventeen participants performed a NHE intervention and had their BFLH scanned using US ten times over a three-week period to track the acute adaptations in muscle architecture to NHE.  Additionally, BFLH muscle architecture was measured using a direct method from the original images in which the full fascicle was visible, and an extrapolation technique from a cropped sample of the original image. Twenty-one males and twenty-one females participated in the sprinting study, which included: US screen, flexibility test, maximum speed test, and maximum strength testing through NHE and an isokinetic dynamometer. The relationships between hamstring injury risk factors were assessed using a bivariate (Pearson) correlation, while the differences between males and females were estimated using paired-sample t-tests. A two-way repeated measures ANOVA was used to compare the muscle load between individual muscles and individual trials, consisting of maximum speed, pre-fatigue, and fatigued trials.

Results and Discussion: The extrapolation of cropped US images led to an overestimation of muscle architecture when compared with the direct measurement of the full FL, with limits of agreement showing an increase beyond the minimal detectable change for muscle thickness (MT) (+1.7 mm) and FL (+6.8 mm), but not for pennation angle (PA) (+0.62°). Wherever possible, direct measurement of the full FL should be made, and improvements in extrapolation methods is warranted. Post NHE intervention, changes were observed for PA (≤4.9° or 33.7%) and FL (≤22.5 mm or 23.2%) but not for MT (≤0.4 mm or 1.8%), which would be considered as a risk factor for HSI. Future research should endeavour to be consistent with muscle architecture reassessment points as variation in the timing could significantly influence their results.

Very strong relationships (R > 0.8) were found between maximum speed and muscle load during the swing phase of sprinting gait, while strong (R > 0.6) relationships involved flexibility and strength parameters. As males typically run faster than females, this may go some way to help explain the increased risk of HSI in males. Nevertheless, it is difficult to determine if differences in muscle load are mainly due to running speed or other physiological differences between the sexes, which themselves would be likely to influence running speed as well. Males showed higher muscle load in BFLH during sprinting and had displayed greater lower body strength than females. Although both sexes were observed to undertake the highest muscle load at maximum speed, BFLH had the highest muscle loads in males while semimembranosus had the highest muscle loads in females.