Central somatosensory changes associated with improved dynamic balance in subjects with anterior cruciate ligament deficiency

Abstract

To examine the mechanisms underlying return to pre-injury function in individuals with anterior cruciate ligament deficiency (ACL-D), we grouped 15 individuals (18–50 years of age) with ACL-D by functional status and strength (i.e. copers, non-copers and adapters) and compared measures of proprioception, somatosensory evoked potentials and neuromuscular responses to dynamic testing between groups. Seven subjects without ACL-D provided a comparative sample for dynamic balance testing (DBT). DBT consisted of bilateral EMG recordings of anterior tibialis, medial gastrocnemius, medial hamstrings and quadriceps during toes-down platform rotation. Relative latencies and relative amplitudes were calculated. Somatosensory evoked potential (SEPs) testing was based on identifying the presence or absence of the P27 potential. Proprioception was tested using threshold to detection of passive movement (TDPM). Those with the highest level of function, the copers, had a proprioceptive deficit, loss of P27 and altered postural synergies consisting of earlier and larger hamstring activation. Conversely, those with the lowest functional status, the non-copers, had strength and proprioception deficits, intact SEPs and inconsistent postural synergies. These results suggest that changes in central sensory representation may facilitate altered postural synergies that enable return to pre-injury functional status.

Introduction

Mechanisms underlying recovery of function in the face of peripheral somatosensory loss and mechanical joint instability due to anterior cruciate ligament rupture are not fully known. Some individuals with anterior cruciate deficiency (ACL-D) return to full contact sports, while others experience “giving way” and instability during activities, such as walking. Research has demonstrated that knee laxity secondary to ACL-D is unrelated to the ability to return to sports [1]. The suggested mechanisms of recovery for this group include: (1) maintenance of proprioception at the knee, as measured by threshold to detection of passive motion [2], [3], (2) activation of a long-loop, capsular-hamstring reflex due to increased mechanical laxity at the joint [4] and (3) changes in central representation of proprioceptive input [5]. These dissimilar conclusions may be explained, at least in part, by the use of different protocol and lack of control for confounding variables, such as strength. It is possible that an interaction of these potential mechanisms underlies recovery of function.

In previous work, we reported [6] an altered muscle activation pattern during gait in individuals with ACL-D without a strength deficit, but with altered somatosensory evoked potentials (SEPs). These findings, although preliminary, support the idea that altered central representation of proprioception contributes to the emergence of compensatory neuromuscular patterns enabling functional recovery. Although investigators agree that individuals with ACL-D present with altered synergies when dynamic balance is challenged [4], [7], [8], [9], reports conflict with regard to the types of changes that occur. A lack of comprehensive testing has prevented identification of the mechanisms involved. In the present study, we examined lower extremity muscle responses to an unexpected dynamic platform tip in individuals with ACL-D, and compared responses between groups of individuals with ACL-D at various functional levels, as well as comparing those with ACL-D to unimpaired individuals.

Di Fabio et al. [4] recorded muscle responses to a horizontal translation of the support surface during bilateral and single leg stance in 12 subjects with unilateral ACL-D. These authors reported that in response to anterior, but not posterior, perturbation in bilateral stance, hamstring activation was faster and less variable in the ACL-deficient limb. Similarly, in response to anterior perturbation during unilateral stance, the ratio of hamstring/quadriceps amplitudes showed a dominance of hamstring activity on the affected limb. These authors [4] concluded that the generalized increased gain of hamstring responses was due to a central change secondary to increased input from the joint capsule of the involved limb. Likewise, Johansson et al. [10] suggested that as a consequence of the ACL injury, there is a mismatch between the ensemble sensory feedback and the existing motor program. The mismatch causes a change in the motor program (altered postural synergies to stabilize the knee). Although several studies [2], [3], [11], [12] have found that proprioception was diminished in ACL-D patients, conclusions that this reduced proprioception is responsible for decreased balance should be guarded. Weakness alone can alter neuromuscular responses as well as tests for proprioception thresholds [13], [14]. To clarify these questions requires concurrent measurement of proprioception, muscle synergies, central changes and functional status, which was not done in these studies. Furthermore, the use of horizontal perturbations to examine the role of proprioception may not be optimal.

Tests of the integrity of sensory input that may trigger muscle activation patterns in this population should enable isolation of somatosensory input, as well as challenge the stability typically provided by the ACL (i.e. anterior tibial translation). The use of an anterior horizontal translation, as used by Di Fabio et al. [4] causes both a relative ankle plantarflexion (somatosensory input) and a backward sway, which is known to trigger vestibulospinal responses [15]. This confounds the interpretation of responses. Beard et al. [16] and others [17] used an unexpected anterior-directed step force to the back of the lower leg during stance with the knee in flexion to examine muscle responses in subjects with ACL-D. This perturbation results in ankle dorsiflexion and an anterior translation of the tibia. Using this protocol, Beard et al. [16] reported increased latency of the hamstring in the affected limb, in contrast to the results reported by Di Fabio et al. The conflicting results may be due to the different biomechanical demands of each protocol, and the addition of vestibular stimulation in the protocol used by Di Fabio et al. Therefore, as noted by Allum and Honegger [18], to optimally examine the response to proprioceptive stimulation at the knee, rotational movement of the platform should be used. This type of test has not been reported in subjects with ACL-D. Despite these limitations, it is apparent that subjects with ACL-D have altered postural synergies. While it has been suggested that central nervous system changes are responsible for these altered synergies, few studies have attempted to measure the central changes [5], [19], [20].

Investigators have determined that somatosensory evoked potentials (SEPs) provide a reliable measure of cortical representation of the integrity of somatosensory input [21]. Therefore, damage of the receptor or pathway could alter the response. SEPs are assessed by stimulating a peripheral neural structure, such as a peripheral nerve and recording cortical potentials from surface electrodes placed on the scalp. Potentials are defined by either P or N, indicating positive or negative polarity and a number, which is an indication of the latency of the potential. The P27 potential is observed with stimulation of the common peroneal nerve (CPN) of either leg in individuals without somatosensory deficits. It is recognized as representative of activity in the primary sensory cortex [21]. Based on SEPs in response to stimulation of the CPN and proprioceptive testing of the knee using TDPM in subjects with ACL-D, Valeriani et al. [5] reported that 53% had proprioceptive loss. Of these, 70% demonstrated loss of the P27 potential on the involved side and retention of the response with stimulation of the uninvolved side. These investigators concluded that loss of the P27 potential represented central reorganization due to proprioception loss resulting from ACL rupture. However, no measures of function were obtained.

Clearly, to clarify the mechanisms underlying return to pre-injury functional status or a failure to do so, in individuals with ACL-D requires measurement of functional status, proprioception, strength and central changes. The purpose of this study was to compare changes in neuromuscular activation in response to perturbation, proprioception as measured by TDPM and central representation of proprioception in individuals with ACL-D with differing levels of functional recovery. We anticipated that the pattern of neuromuscular responses would be unique and that central somatosensory changes would be evident in those functioning at the highest level.