Submaximal Force Steadiness and Accuracy in Patients With Chronic Ankle Instability

Participants

A total of 42 physically active college students, consisting of 21 participants with CAI and 21 healthy control participants, were recruited from a university population. Using an a priori power analysis and previous data (isometric force steadiness) with α, β, and Cohen d values of .05, 0.2, and 0.93, respectively,⁸ we determined that a sample size of 36 participants was needed. Exclusion criteria were a history of lower limb surgery or fracture, or neurologic disorder in the participant's lifetime or any sport-related injuries in the previous 3 months. Participant demographic information is shown in Table 1.

Table 1 Participant Demographics

View Fullscreen

Recruits with CAI were identified based on the International Ankle Consortium guidelines.²² We used self-reported disability questionnaires: the Foot and Ankle Ability Measure for Activities of Daily Living (FAAM–ADL), Foot and Ankle Ability Measure for Sports (FAAM–Sports), and Ankle Instability Instrument (AII). Specific inclusion criteria for participants with CAI were (1) a history of ankle sprain that occurred at least 3 months before data collection; (2) a score of <90% on the FAAM–ADL; (3) a score of <80% on the FAAM–Sports; (4) at least 5 yes answers, including to question 1, on the AII; and (5) a history of physical activity performed at least 3 d/wk, for a total of 90 min/wk, in the previous 3 months.²⁰ Specific inclusion criteria for healthy control individuals were (1) no previous ankle sprain, (2) a score of 100% on the FAAM–ADL, (3) a score of 100% on the FAAM–Sports, (4) zero yes answers on the AII, and (5) a history of physical activity performed at least 3 d/wk, for a total of 90 min/wk, in the previous 3 months.²⁰ All individuals provided informed consent, and the study was approved by the appropriate institutional review board.

Experimental Procedures

The experimental procedures are illustrated in Figure 1. A Biodex dynamometer and Advantage Software (model 3 dynamometer; Biodex Medical Systems, Inc) were used to measure the maximal voluntary isometric contraction (MVIC) and force steadiness and accuracy of the ankle evertors and invertors and the hip abductors. We measured the MVIC to permit comparisons of the maximal force of those muscles between groups. Participants were provided an opportunity to become familiar with the isokinetic dynamometer and testing procedure and to perform as many warm-up repetitions as desired (at least 5). During the practice session for the MVIC measures, they were instructed to perform the task at various force outputs (25%, 50%, 75%, and 100% of MVIC). After a 3-minute rest, participants performed 3 MVIC trials by contracting the muscles (ankle evertors and invertors and hip abductors) as hard as possible for 3 seconds while minimizing other movements. A 1-minute rest was allowed between trials. Previous authors^11–15 examined various force increments (10%, 20%, 25%, 30%, 40%, and 50% of MVIC) to measure force steadiness in different joints. Because no earlier investigators measured force steadiness in patients with CAI, we chose to use 10% and 30% of each person's MVIC based on a study⁶ in which the authors measured force sense for the same motions. Participants were informed of how force steadiness and accuracy would be measured. During the test, they were instructed to attempt to stay as close as possible to the target force (10% or 30% of MVIC) for 15 seconds. Five trials were performed for each muscle; the first 2 were considered practice trials. Participants were able to adjust their force using a monitor (1 m away) that showed the actual force. While participants performed the force-steadiness trials, 1 examiner indicated the target force line, so that each person easily recognized the target forces. The examiner randomly assigned a target force, 10% or 30% of MVIC, to each person.

View Figure

• Download as Powerpoint
• Download Figure

Participant Position on the Dynamometer

All participants wore athletic shoes (model T-Lite XI; Nike, Inc) during testing. The shoes were tightly secured to the footplate to minimize unnecessary movement between the shoe sole and footplate surface during the ankle measurements (Figure 2A). For the hip measurements, participants were positioned on their side in the dynamometer chair facing the dynamometer, thereby allowing less space for compensatory movements (Figure 2B). This position provided a backrest and allowed each person to use the handhold in front, thereby minimizing trunk and pelvic rotation during testing. The position was referenced using the Biodex System 3 application and operation manual.

View Figure

• Download as Powerpoint
• Download Figure

Data Processing and Statistical Analysis

The data for all 3 MVIC trials were processed using custom-written MATLAB software (The MathWorks, Inc) to determine maximal torque. We also used the MATLAB software to analyze submaximal force steadiness and accuracy. For all conditions, the first 2 trials were discarded and the central 10 seconds of data (20%–87% of the total time) were assessed for the remaining 3 trials (Figure 3). Force steadiness and accuracy were obtained following the methods of an earlier study.⁸ Specifically, the SD across the 10 seconds of data provided the participant's response signal without reference to the target force level, which was force steadiness. We calculated force accuracy as the root mean square of the difference between the participant-generated force and the target force. The MVIC was analyzed using an independent t test for group effect (CAI versus healthy control). Force steadiness and accuracy were evaluated using a 2-way (2 groups [CAI and healthy control] × 3 motions [ankle eversion and inversion and hip abduction]) analysis of variance for the group effect, motion effect, and group × motion interaction. We conducted Tukey honestly significant difference (HSD) post hoc tests for pairwise comparisons. The experiment-wise type I error rate for all tests was set at P < .05. Cohen d effect sizes were calculated to provide the magnitude of differences between groups (from 0.21 to 0.5, small; 0.51 to 0.8, moderate; >0.8, large).

View Figure

• Download as Powerpoint
• Download Figure

Force Steadiness (SD)

The force-steadiness data at 10% and 30% of MVIC for ankle eversion and inversion and hip abduction for both the CAI and control groups are shown in Table 2. No interaction was present between group and muscle at either 10% or 30% of MVIC (F_5,120 = 1.42, P = .25, and F_5,120 = 1.32, P = .27, respectively). A group main effect was noted at 10% of MVIC for force steadiness (F_5,120 = 12.67, P < .001) but not at 30% of MVIC (F_5,120 = 1.76, P = .19). We observed motion main effects at 10% and 30% (F_5,120 = 136.09, P < .0001, and F_5,120 = 100.77, P < .0001, respectively) of MVIC. A Tukey HSD post hoc test indicated that the hip abductors demonstrated less steadiness than the ankle evertors and invertors at 10% and 30% of MVIC (P < .0001). Even though no interaction was present between group and muscle, the Cohen d showed a large effect size (1.18) between groups at 10% of MVIC in inversion. In summary, patients with CAI displayed less steadiness in inversion than the control group.

Table 2 Force-Steadiness and Force-Accuracy Errors

View Fullscreen

Force Accuracy (Root Mean Square)

The force accuracy error data at 10% and 30% of MVIC for ankle eversion and inversion and hip abduction for both the CAI and control groups appear in Table 2. Group × muscle interactions were noted at both 10% and 30% of MVIC (F_5,120 = 7.57, P = .01, and F_5,120 = 4.22, P = .02, respectively). A Tukey HSD post hoc test indicated that the CAI group showed less inversion accuracy than the control group at 10% of MVIC (P < .001). There was a group main effect at 10% of MVIC in force accuracy (F_5,120 = 7.42, P < .001), but no effect at 30% of MVIC (F_5,120 = 3.74, P = .06). Motion main effects were evident at 10% and 30% of MVIC (F_5,120 = 25.08, P < .0001, and F_5,120 = 15.46, P < .0001, respectively). A Tukey HSD post hoc test indicated that the hip abductors showed less accuracy than evertors and invertors at 10% (P < .0001) and 30% of MVIC (P < .0001). In addition, the Cohen d showed a medium effect size (0.65) between groups at 10% MVIC for eversion. In summary, the patients with CAI exhibited less accuracy in eversion and inversion compared with the control individuals.

Maximal Voluntary Isometric Contraction

The MVIC data for ankle eversion and inversion and hip abduction for both the CAI and control groups appear in Table 3. No group difference was present for MVIC in eversion and inversion (t_1,124 = 2.25, P = .14, and t_1,124 = 1.39, P = .24, respectively), but participants with CAI showed less force output in hip abduction than control participants (t_1,124 = 58.06, P < .0001).

Table 3 Maximal Voluntary Isometric Contractions

View Fullscreen

Differences Between Groups in Force Steadiness and Accuracy

Our participants with CAI maintained less force steadiness in their ankle invertors during isometric contractions. In addition, they were less accurate in ankle eversion and inversion. The increased variability in force steadiness and accuracy represents altered motor-unit recruitment and firing rates, impaired proprioceptive information, increased activation of synergist and antagonist muscles, and altered spinal interneuron modulation of motor-neuron firing.⁸ We propose that the observed alterations in force steadiness and accuracy may be due to impaired proprioceptive function in these patients. Proprioceptive sensory inputs from muscles, tendons, and ligaments are transmitted from the peripheral nervous system to the CNS; this process is necessary for appropriate neuromuscular control.²³ Neuromuscular control can be affected by (1) the collection of less peripheral information because of damaged proprioception, (2) an inability to integrate the peripheral information in the CNS, or (3) an inability to send out the centrally mediated information to the motor units.⁹ Thus, our results may be attributed to one or more of the aforementioned factors. Moreover, Chung-Hoon et al²⁴ explained that less steadiness in force output may be caused by presynaptic inhibition of Ia afferents. Because of the depolarization of primary afferent fibers by interneurons, the input from Ia afferents to the active motor-neuron pool may be inhibited and consequently affect motor-unit activation in maintaining a certain force.²⁵ Accordingly, patients with CAI may be unable to regulate presynaptic inhibition compared with healthy control individuals.²⁵ Furthermore, Docherty and Arnold⁷ suggested a significant relationship between ankle instability and force sense. They showed that patients with functional ankle instability had deficits in precise force sense and joint position sense.⁷ Because CAI is associated with impaired proprioception, strength, and postural control as a result of repeated LASs,² reduced force accuracy could be a consequence of the injury.

Compared with the control group, our CAI group demonstrated less steadiness and accuracy in the ankle invertors but not the evertors. A possible explanation is that the invertors had deficits in motor control and more degrees of freedom in movement during force generation than the evertors. Specifically, previous authors^2,18 reported deficits in motor control such as neuromuscular inhibition and muscle weakness not only in the evertors but also in the invertors. Inversion has a larger range of motion and involves more muscle fibers for force generation than eversion. Thus, the invertors may display less steadiness and accuracy because inversion has a greater range of motion to generate force than eversion. Based on the literature, no direct relationship exists between force control and the degrees of freedom in force generation. However, absolute error was compared among different joints in the lower extremity during force steadiness in several investigations.^13–15 The earlier results combined with ours support the idea that the greater the lower extremity joint during force generation, the greater the error during force control (error increases from eversion to inversion to hip abduction to knee extension). Both impaired motor control and the greater degrees of freedom in the invertors led to impaired force steadiness and accuracy.

A Novel Measurement for Patients With CAI: Force Steadiness and Accuracy

Unlike previous authors^6,7 who measured force steadiness and accuracy, we allowed our participants to view their force steadiness and accuracy output on a monitor and adjust their target force to approach the target forces. Earlier investigators demonstrated that visual feedback enhanced not only isokinetic muscle force²⁶ but also neuromuscular control, such as interlimb coordination.²⁷ However, we found that the CAI group had more difficulty than the control group maintaining target forces. Thus, patients with CAI may have difficulty reducing force error. In other words, their ability to integrate sensory information in the CNS to produce the appropriate motor output may be impaired. To explore this idea, more work is needed to examine the effect of visual information on force steadiness and accuracy. Practically, deficits in the ability to integrate sensory information could easily result in alterations in movement and loads and ultimately increase the injury risk.²⁸

Changes in Neuromuscular Control in the Proximal Joint

Researchers^19,21 have studied hip-joint neuromuscular alterations in those with LAS and CAI. In those with CAI, hip biomechanics were altered, which could have been caused by deficits in neuromuscular control.²¹ Decreased hip strength also influenced dynamic postural control, which needs appropriate neuromuscular control in the CAI population.¹⁹ In accord with previous studies,^19,21 our results indicated that the CAI group demonstrated a smaller MVIC in hip abduction than the control group. However, submaximal force steadiness and accuracy did not differ between the groups. The task in this study was submaximal force steadiness, which is less dynamic than the jump-landing task and Star Excursion Balance Test.^20,21 This less demanding task may be insufficient to induce differences in force steadiness and accuracy. Additionally, we measured isometric hip-abductor force. If we were to measure concentric or eccentric force, which is a more dynamic task, we might elicit a difference between conditions. Thus, future work is needed to measure force steadiness and accuracy during concentric or eccentric contractions.

Differences Between Muscle Groups in Force Steadiness and Accuracy

The hip abductors showed more errors in force steadiness and accuracy than did the ankle evertors and invertors. This finding suggests that as more muscle fibers were recruited, participants' steadiness decreased. Because limited data are available regarding force steadiness in large and small muscle groups, we based our explanation on earlier investigations.^14,15,29 The mass of the knee extensors is greater than that of the hip abductors, and the mass of the hip abductors is greater than that of the ankle evertors or invertors.²⁹ Our absolute error at 10% of MVIC force steadiness for ankle eversion and inversion was less than 0.4 N/kg. However, absolute errors of force steadiness for the quadriceps in patients with knee¹⁵ or hip osteoarthritis¹⁴ were higher than 1.3 N/kg (2.13 ± 1.51 and 1.3 ± 0.94 N/kg, respectively). Therefore, we propose that the number of muscle fibers recruited may affect the ability to maintain submaximal force steadiness. A relationship between the number of muscle fibers and force steadiness and accuracy would be a valuable area of future examination.

Clinical Implications

Our findings of impaired ankle and hip force steadiness and accuracy in patients with CAI provide useful insights for clinicians developing rehabilitation protocols. An indirect indication was that the CAI group had proprioceptive deficits in the force steadiness and accuracy of ankle eversion and inversion. This impaired proprioception might lead these patients to be more susceptible to injury positions as they have difficulty integrating the peripheral information and correcting their movement in relation to visual information. Restoring proprioceptive function of the ankle and hip plus visual training may be key to improving clinical outcomes for this population.^21,30 Our data suggest that clinicians should continue to focus on restoring proprioceptive function in both the distal and proximal joints in conjunction with visual information to improve force control in various movements. Moreover, movement-related functional rehabilitation exercises are necessary for adjusting and correcting movement errors and increasing the ability to produce appropriate force during a given task.

Limitations

We only measured an isometric contraction of the ankle evertors and invertors and the hip abductors. This measure may have been insufficient to fully understand the somatosensory function of the involved joints during sports or activities of daily living. Although errors in force steadiness and accuracy during concentric and eccentric contractions of the knee extensor were not different between concentric and eccentric conditions,⁸ the ankle musculature may show different results.