Compensatory Load Redistribution in Forelimb and Hindlimb Lameness

At the walk, in unilateral forelimb and hindlimb lameness, reduced loading of the lame limb is primarily compensated by concurrently loaded limbs. At the trot, the horse's attempt to cope with weight-bearing forelimb lameness causes a reduction of stride impulse and its redistribution to the contralateral forelimb and diagonal hindlimb without producing an equivalent, compensatory increase of peak vertical forces in these limbs. In hindlimb lameness, impulse redistribution is limited to the contralateral hindlimb without compensatory increase in peak vertical forces.

1. Introduction

Kinetic analysis studies the forces that are responsible for a movement. Human sensory perception is not able to derive kinetic information from visual gait assessment.

The first kinetic studies on equine locomotion date back to 1874. Etienne Jules Marey's (1830 - 1904) Chaussure exploratrice, which consisted essentially of an India rubber ball filled with horse hairs, was fitted between the branches of the horseshoe over the frog; this device recorded pressure changes at hoof placement. The pressure fluctuations were transmitted by airtight rubber tubes and registered on a charcoal-blackened rotating cylinder that the rider held in his hand. Marey was the first to refer to vertical forces when describing equine stride patterns.

Kinetics distinguish between internal and external forces and torques. External forces refer to forces that are generated from outside a system, such as the force that the ground exerts on the body or the so-called ground reaction force (GRF). Internal forces, such as tendon forces or bone strains, can be determined directly; however, these measurements are invasive and therefore, limited to research applications. A non-invasive method of estimating internal forces and torques is to make use of inverse dynamic models. Inverse dynamics calculate the internal forces by applying the laws of dynamics (Isaac Newton, England, 1642 - 1727) to a linked multi-segment model (Joseph Louis Lagrange, France and Italy, 1736 - 1813) of the respective object of study. The basis for these calculations are time-locked kinetic and kinematic data. Radiography of the specific anatomical region supplies the geometric information (e.g., moment arm of the tendons at the coffin and fetlock joints).

2. Materials and Methods

External forces can be measured directly by the use of a force plate [1,2], force shoes [3-7], or more sophisticated force measuring systems such as the Equine Gait Analysis (EGA) System [8,9]. The most established technique is the force plate [a-c]. The other concepts are mainly reserved for research investigations. Standard force plates are able to split the input force into its three orthogonal components: the transverse-horizontal force (Fx), the longitudinal-horizontal force (Fy), and the vertical force (Fz; Fig. 1).

Figure 1. Forelimb ground reaction forces (GRF) of a horse (A) at the walk and (B) at the trot. The Fz traces are typically biphasic at the walk with a higher second peak in the forelimbs and a higher first peak in the hindlimbs. At the trot, canter, and gallop, the Fz curves show an impact force peak 30-35 milliseconds after touch-down and a single force maximum around midstance. Fy traces are characterized by negative braking forces in the first half of stance followed by positive propulsive forces. Fx are very small in magnitude--only 2% body weight (bwt) at the trot. Fz, vertical forces (light grey line); Fy, longitudinal forces (grey line); Fx, transverse forces (black line).

Force plate reference data exist for sound horses walking [10,11], sound horses trotting [12,13], sound horses cantering [14], horses with riders walking, horses with riders trotting [15], horses jumping a 0.8 - 1.3 m high fence, and horses landing from jumps [16]. At the walk (Fig. 1A), peak Fz (Fzpeak) reaches an average of 66% (second peak) of body weight (bwt) in the forelimbs and 51% (first peak) in the hindlimbs [10]. At the trot (Fig. 1B), Fzpeak reaches an average of 118% in the forelimbs and 104% in the hindlimbs [12]. At the canter, Fzpeak reaches an average of 101% in the trailing hindlimb, 115% in the leading hindlimb, 147% in the trailing forelimb, and 122% in the leading forelimb [14]. When clearing a 0.8-m high fence, vertical GRF amplitudes are similar in magnitude to those measured at the canter. Data of a single horse indicated that peak forces increased sharply with greater fence heights. Jumping a 1.3-m high fence at the right-hand canter, the highest Fzpeak at take-off is observed in the trailing forelimb and hindlimb (173% and 143% bwt, respectively); during landing, Fzpeak decreases gradually along the step sequence from 204% bwt in the trailing forelimb to 153% in the leading forelimb and trailing hindlimb and 122% in the leading hindlimb [16]. Particularly remarkable is the observation that the magnitude of forces varies according to the jumping technique of the specific horse. Ratzlaff et al. [17] measured Fz in horses at racing speed (13.7 - 15.8 m/s) using instrumented horseshoes. On the straight, the greatest Fz was exerted on the leading forelimb, followed by the leading hindlimb; on banked turns, the greatest Fz was taken by the leading and the trailing forelimbs, confirming the observation that the majority of racing injuries occur in the forelimbs.

When using force plates, problems in obtaining repeatable, constant speed trials or targeting the platform may present substantial restrictions, especially when dealing with quadrupeds. Depending on the gait, 2 - 20 attempts are needed to register a single, valid foot strike [10,12,14]. Furthermore, with a single force plate, GRF data of only one limb can be recorded at a time while at least one other limb supports the body concurrently. To get a complete indication on the load redistribution in case of lameness, force-time histories of all four limbs are indispensable.

A treadmill equipped with a dynamometric platform combines the benefits of kinetic analysis with the advantages of the treadmill, such as movement velocity control and data acquisition over multiple consecutive motion cycles of all four limbs simultaneously [18]. Controlling velocity of movement is a prerequisite for accurate gait analysis, because kinematic and kinetic parameters are known to be velocity dependent [19-22]. The instrumented treadmill is fully operational in a clinical set-up. Typically, up to 50 strides per limb, at both the walk and the trot, are recorded in <5 min. The calculation of the Fz traces starts during data acquisition, and force curves are presented in real time on a computer screen. Fourteen force, eleven temporal, and six spatial parameters are derived from the raw data and presented graphically a few seconds after completion of the measurement (Fig. 2). This allows, for example, the instantaneous assessment of diagnostic blocks.

Figure 2. Force and time parameters extracted from the force traces at the trot. FL, left forelimb; FR, right forelimb; HL, left hindlimb; HR, right hindlimb; Fzpeak, vertical force peak; TFzpeak, time of vertical force peak; Iz, vertical impulse; SD, stride duration; StD, stance duration; SwD, swing duration; StpDcl, contralateral step duration; StpDil, ipsilateral step duration; SpD, suspension duration; TAP, time of advanced placement; TAC, time of advanced completion.

Kinetics of lameness have been studied by using equine patients with clearly diagnosed lameness (navicular disease or tendonitis of the superficial digital flexor tendon) [9,23-29] or by applying lameness models. Intra-articular injection of endotoxins was used to induce synovitis of the distal tarsal joints [30], and a chip fracture of the radial carpal bone was created surgically in an osteoarthritis lameness model [31,32]. These models were used to induce a mixed, supporting-swing limb lameness. Intratendinous injection of collagenase was used to induce superficial digital flexor tendonitis [33-36]. All methods resulted in a relatively long-standing lameness, sometimes without return to normal function. Another method, which imitates the effect of a stone trapped under the horseshoe, was described by Merkens and Schamhardt [37] and applied in several studies [38-40]. Bolts are screwed into nuts welded to the inner rim of each branch of the horseshoe. When the limb is loaded, pressure is applied to the corium of the sole. By tightening or loosening the bolts, various degrees of lameness can be elicited. Additionally, the duration of the lameness can easily be controlled, and the lameness is fully reversible.

3. Results

From force plate studies, it is known that weight-bearing lameness mainly affects Fz and Fy, whereas changes in Fx are negligible. At the walk, in unilateral forelimb lameness as well as hindlimb lameness, reduced loading of the lame limb is primarily compensated by the contralateral limb and to a lesser extent, by concurrently loaded limbs [38].

At the trot, four compensatory mechanisms serve to reduce structural stress (i.e., Fzpeak on the affected limb [Fig. 3]) [39,40].

Figure 3. Changes in Fzpeak with increasing (A) FL lameness and (B) HL lameness in horses at the trot. Data are presented as mean ± SEM. The percentage information represents the proportion of the sound value. Sound, condition before inducing unilateral lameness; 1, subtle lameness; 2, mild lameness; 3, moderate lameness; a, significant difference (p < 0.05) compared with the sound condition; b, significant difference (p < 0.05) compared with the preceding condition.

1. With increasing lameness, horses reduce the total vertical impulse per stride (IzSD) by increasing stride frequency (SF; Table 1). The momentum theorem explains the interdependence of impulse and the time during which the forces act (FL, left forelimb; FR, right forelimb; HL, left hindlimb; HR, right hindlimb):

IzSD=IzFL+IzFR+IzHL+IzHR=bm/g/SD

During a complete stride cycle, the time-integrated vertical GRFs (IzFL + IzFR + IzHL + IzHR) equal the gravitational force throughout stride duration (bm, body mass of the subject; g, gravitational acceleration; SD, stride duration).

2. The diagonal vertical impulse (Izdiag) decreases selectively in the lame diagonal, which causes a shortened suspension phase (SpD; Table 1) and a faster transition from the lame to the sound diagonal stance (contralateral step duration, StpDcl; Fig. 4) [41]. In correspondence, a reduction in vertical trunk movement during the lame diagonal stance and an increase in affected vertical trunk movements during the sound diagonal stance can be observed [42].

Figure 4. Changes in StpDcl with increasing (A) FL lameness and (B) HL lameness in horses at the trot. FL > FR, transition from FL to FR; FR > FL, transition from FR to FL; HL > HR, transition from HL to HR; HR > HL, transition from HR to HL.

3. In forelimb lameness, the impulse (Iz) is shifted within the lame diagonal to the hindlimb and during the opposite diagonal stance to the forelimb (Fig. 5A) [32,43]. In hindlimb lameness, the impulse shifts occur predominantly from the affected to the contralateral hindlimb (Fig. 5B).

Figure 5. Changes in Iz with increasing (A) FL lameness and (B) HL lameness in horses at the trot.

The weight-shifting mechanism along the longitudinal axis of the horse corresponds with both the changes in body center of mass movement [44] and the conclusions drawn from the modeling of compensatory head movements in lame horses [45]. The reduction or even suppression of the downward head acceleration during the lame diagonal stance phase decreases the momentum in the trunk; this unloads the lame forelimb, but it also increases the loading of the diagonal hindlimb. During the sound diagonal stance, the distinct vertical head nod together with the higher horizontal braking forces in the contralateral forelimb produce a momentum in the trunk, which increases the loading of the contralateral forelimb and decreases that of the ipsilateral hindlimb [32,33]. The model calculations estimated that a difference of only 10 cm in the vertical amplitude of the head during the stance phases of the lame and sound forelimb causes differences in the Fz of nearly 500 N and differences in the sagittal torque acting on the trunk of ~230 Nm [45].

4. The rate of loading and peak forces are reduced by prolonging the stance duration (StD) in the affected limb as well as in the contralateral limb (Fig. 6).

Shortening of StD is often empirically claimed to be an acoustic indicator for weight-bearing lameness. Because the end of stance is silent, StD is not assessable acoustically; additionally, the bilateral equal prolongation of StD refutes this theory. However, the asymmetric shortening of contralateral step duration (StpDcl), and therefore, the shorter transition interval from lame to sound limb impact reinforced by the louder impact of the sound limb, is the more plausible lead for acoustic detection of lameness.

Figure 6. Changes in StD with increasing (A) FL lameness and (B) HL lameness in horses at the trot.

The consequences are remarkable: with these load shifting mechanisms, the relief of peak forces in the lame limb is not only effective, but it even suppresses an equivalent compensatory overload situation in the other limbs. Except in moderate forelimb lameness, where Fzpeak increase slightly in the diagonal hindlimb, no compensatory increase in Fzpeak can be observed in other limbs (Fig. 3A) [39]. Likewise, in mild and moderate hindlimb lameness, no compensatory increase in Fzpeak can be observed in other limbs (Fig. 3B) [30,40]. Forelimb lameness affects Fy as well. Morris and Seeherman [32] observed decreased braking forces in the lame forelimb and increased braking forces in the contralateral limb. Additionally, they observed that the propulsive forces decreased in the ipsilateral hindlimb compared with the diagonal hindlimb [32].

4. Discussion

The analysis of GRF is a reliable method to quantify lameness. The force parameters Fzpeak and Iz proved to have the highest limb specificity and sensitivity to grade lameness. Temporal stride variables, when considered on their own, are of questionable value in detecting lameness. First, mild lamenesses do not show significant temporal deviations from the sound stride pattern. Second, key parameters, such as StD or the time of diagonal advanced placement, maintain their left-to-right symmetry with increasing lameness [39-41,43,46]. Temporal asymmetry is better interpreted as a sign of individual locomotor pattern, also known as sidedness, handedness, or laterality [47,48]. However, temporal parameters characterizing the airborne phase of the trot (StpDcl or SpD) change asymmetrically with increasing lameness [39,-41]. Controlling velocity of movement is a prerequisite for accurate gait analysis, because kinematic and kinetic parameters are known to be velocity dependent. Recording single strides with a force plate may be very time consuming; this limits the use of this method in a clinical set-up.

Horses redistribute load using a consistent strategy to compensate for pain in a forelimb and hindlimb without causing an overload situation in other limbs. However, at higher speed (i.e., at higher stride frequencies and therefore, shorter stance and swing durations), the possibility of adapting the inter- and intra-limb timing is limited; therefore, compensatory overload in the other limbs cannot be ruled out. In more severe lameness conditions, the horse will be forced to change to another compensatory mechanism entirely, reducing StD of the affected limb. This will obviously disturb the continuous cadence of the gait and may induce compensatory overload in the diagonal and contralateral limbs.

Footnotes

[a] AMTI, Watertown, MA 02172.

[b] Bertec Corporation, Columbus, OH 43229.

[c] Kistler Instrument Corp., Buffalo, NY 14228.