Motor control and strength as predictors of hamstring injury in elite players of Australian football

Abstract

Objectives. To investigate motor control as a possible contributing factor in hamstring injuries using a test movement similar to the injury-critical phase during running. Leg swing movement discrimination (MD), thigh muscle strength as tested by Cybex isokinetic dynamometry and previous hamstring injury history were all assessed to determine any association with subsequent hamstring injury.

Design. Prospective and retrospective observational and analytical cohort study.

Setting. A professional football club and a university laboratory.

Participants. Twenty elite level players of Australian football.

Results. In the 2 years prior to testing, 7 subjects had sustained a hamstring injury. At the time of testing, there were no significant differences between subjects never injured and those previously injured. In the two years following testing, 6 players sustained a hamstring injury. Mean MD score and concentric hamstring-to-quadriceps strength ratio of subsequently injured subjects were significantly worse than those uninjured (p<0.05). Optimum cutoffs on these measures were determined.

Conclusion. Both a lower MD ability score for the backward swinging leg and an imbalance of thigh muscle strength were predictive of hamstring injury.

Introduction

Hamstring strain injuries are common in all football codes and sports involving sprinting, and are the most frequently occurring and recurring of all injuries in Australian football. At the elite level, hamstring injuries occur at a rate of 6.2 injuries per club per season, and result in 21.2 missed player games per club per season (Orchard and Seward, 2002). These rates are the highest of all the elite level football codes in Australia (Seward et al., 1993). Hamstring injuries have the highest recurrence rate of all football injuries and notwithstanding the best rehabilitation attempts, more than one in three (34%) injuries recur within the same season (Orchard and Seward, 2002). Risk has been shown to be increased following previous hamstring injury (Bennell et al., 1998, Garrett, 1996, Orchard, 2001), calf strain injury (Orchard, 2001), and serious knee and groin injuries (Verral et al., 2001). Risk has also been shown to be increased in players of Australian football older than 23 years of age (Orchard, 2001).

Despite the magnitude of this problem in Australian football, the aetiology of such injuries is unclear. Factors that have been suggested to predispose an athlete to hamstring muscle strain injury include muscle weakness, muscle imbalance, poor flexibility, fatigue, inadequate warm-up, poor neuromuscular control and poor running technique (Agre, 1985). There is little evidence for poor flexibility as a hamstring injury predictor (Bennell et al., 1999), and apart from muscle weakness there is no empirical support for any of the other suggested factors (Orchard, 2001).

The hamstring muscle group reaches peak elongation and acts eccentrically at the hip and knee during the late swing phase of the running cycle (Frigo et al., 1979, Simonsen et al., 1985). Kinetic and EMG studies reveal that the hamstrings are most active and develop the greatest torques at the hip and knee during late swing through to the mid-stance phase of running (Mann and Sprague, 1980, Montgomery et al., 1994). It is during these parts of the running cycle that the hamstrings are under the greatest demands and injury most likely (Mann and Sprague, 1980, Stanton and Purdam, 1989). Given the high forces involved, it would seem that hamstring weakness might predispose an athlete to injury, however, to date, there have not been adequate findings to support either hamstring muscle weakness or hamstring-quadriceps strength imbalance as risk factors (Orchard, 2001).

Muscle weakness has been the most extensively investigated of all proposed predisposing factors for hamstring injury. Retrospective studies have suggested a relationship between muscle weakness and hamstring injury (Crosier and Crielaard, 2000, Crosier et al., 2002, Heiser et al., 1984), however, retrospectivity can confound the causes and effects of injury. The literature also contains a small number of prospective studies, which allow firmer conclusions to be drawn regarding the factors involved in the prediction of injury, but the results to date are inconsistent. Two studies using isometric hamstring strength assessment demonstrated an association between injury and a side-to-side deficit of 10% (Burkett, 1970) and a lower hamstring-to-quadriceps ratio (Yamamoto, 1993). The latter finding was supported by a study using isokinetic dynamometry at 60 °/s, which linked injury to a hamstring-to-quadriceps ratio below 0.60 (Orchard et al., 1997). In contrast, Liemohn (1978) did not find any relationship between injury and isometric hamstring-to-quadriceps strength ratio, nor did a larger study examining concentric and eccentric isokinetic strength reveal any association between injury and similar strength ratios (Bennell et al., 1998).

One possible causative factor of hamstring injury, that has not been previously examined, is the accuracy of neuromuscular control of the leg during running. Throughout the running cycle there are many challenging neuromuscular events in which the hamstring acts, for example, to control hip and knee motion in late swing and to provide hip extensor torque in early stance. During sprinting, these events occur over a very short period of time, and if the control and coordination are inadequate, then muscle strain injury may result (Agre, 1985, Bennell et al., 1999). Control of a limb requires that information be obtained and integrated from proprioceptors of the entire limb, and that control will be influenced in part by activity in the opposite limb. Joint or single segment proprioception has been assessed by various techniques that employ joint position testing, kinesthaesia testing or sense of effort testing, often in non-weight-bearing postures (Lephart and Fu, 2000). Waddington and Adams (1999) assessed movement discrimination (MD) at the ankle based on functional movement principles, whereby subjects performed ankle inversion movements in standing and made judgments regarding the extent of these movements without visual input about the movement. This requires the processing of both afferent and efferent information about the lower limb being tested, and performance reflects a subject's ability to do this accurately. An association was demonstrated between poor discrimination ability and previous ankle sprain injury. This task was later extended to enable assessment of MD of the knee during weight-bearing flexion (Waddington et al., 2000).

Utilising the same functional movement principles for the current study on hamstring function, a similar apparatus may be employed to assess the discrimination of movement extent using the backward swinging leg, whilst weight-bearing on the other side, in order to create a functional movement as close as possible to the movement at injury. The position of the player while being tested is selected to recreate the proposed movement region in which hamstring injury occurs: between the late swing to mid-stance phases of the running cycle (Fig. 1). Ability to accurately discriminate swinging leg movements can be seen as reflecting use of proprioception from the lower limbs, and is integral to the motor control of the leg in this region of the running cycle where hamstring injury is likely to occur.

The aim of this study was to assess swinging leg MD, isokinetic hamstring and quadriceps strength, and history of previous hamstring injury, in order to determine any association these factors have with respect to subsequent hamstring injuries in a group of elite Australian football players.