Ankle Ligament Laxity and Stiffness in Chronic Ankle Instability

Methods

Participants

An a priori power analysis was completed (Version 3.1.5, G*Power, Kiel, Germany) from tabled data in similar studies utilizing an instrumented arthrometer to perform an inversion laxity test. For a power of .80 with α ≤ .05, the necessary sample size for the chronic ankle instability–control comparison was n = 10 per group ¹⁹ and for the chronic ankle instability–coper comparison was n = 16 per group. ³⁰ Another study indicated that the necessary sample size was 13 to 16 per group. ⁹ The control-coper comparison would require n = 640 per group. ³⁰ Because this comparison was of least interest, and with the excessive sample size, we used n = 16 as the sample size.

Institutional review board approval was granted, and informed consent was obtained from all participants prior to testing. Participants were 48 volunteers from the community, aged 18 to 30 years, who completed at least 1.5 hours of physical activity per week and were recruited and tested from June 2011 to November 2012 (Table 1). Inclusion criteria for the chronic ankle instability group included a history of moderate-severe ankle sprain ⁸ requiring 3 or more days of partial weightbearing or nonweightbearing, inflammatory symptoms (pain, swelling), disruption of desired physical activities, “giving way” at the ankle with activity, 2 or more episodes of giving way at the ankle in the last 12 months, and a Cumberland Ankle Instability Tool (CAIT) score of 26 or less, indicating decreased function. ¹⁴ In individuals who indicated bilateral instability, the limb with the lower CAIT score was used. Inclusion criteria for copers included a history of moderate-severe ankle sprain requiring 3 or more days of partial weightbearing or nonweightbearing, no complaints of “giving way” of the ankle with activity, 1 or fewer episodes of giving way at the ankle in the last 12 months, and a CAIT score of 28 or greater, indicating good function.^14,38 Controls had no history of lateral ankle sprain, no complaints of giving way, and CAIT scores of 29 to 30, indicating no loss of function. ⁸ Participants were matched between groups by sex, height, and weight (within 10%) and limb dominance. The dominant leg was defined as the limb used to perform 2 of the 3 following tasks: step up, kick a ball for distance, and step forward after a push on the back. ¹⁷ Participants were not matched by age but were all within the range listed and met physical activity criteria.

OPEN IN VIEWER

Table 1.

Participant Demographics and Injury History With 1-Way Analysis of Variance Results.

Variable	Chronic Ankle Instability	Coper	Control	P Value
Sex, n	16 (12 M, 4 F)	16 (12 M, 4 F)	16 (12 M, 4 F)	NT
Mean age, y	21 ± 2	22 ± 4	21 ± 1	NS
Mean height, cm	174 ± 9	175 ± 11	172 ± 8	NS
Mean weight, kg	76 ± 14	75 ± 13	72 ± 11	NS
Mean CAIT score	19 ± 6 a	29 ± 1	30 ± 1	<.001
No. of sprains reported, mean (range)	6 ± 5 (2-21)	2 ± 1 (1-2)	0	NT

Abbreviations: CAIT, Cumberland Ankle Instability Tool; F, female; M, male; NS, not significant; NT, not tested.

a

Chronic ankle instability group is significantly lower than coper and control groups (P < .05).

Exclusion criteria for all groups were vestibular disorder, hereditary nerve or connective tissue disorder, pregnancy, fracture of the lower extremity, previous surgery of the lower extremity, and acute signs of lower extremity injury (pain, heat, redness, discoloration, and swelling). ³⁸

Procedures

After providing consent, participants completed the CAIT and an ankle injury history questionnaire to determine their status on inclusion/exclusion criteria. Demographic information was recorded. Generalized joint laxity was not recorded, as it is not related to ankle joint laxity. ³³ Participants underwent an inversion test using an instrumented ankle arthrometer (LigMaster Version 1.26, Sport Tech Inc, Charlottesville, Virginia, USA) operated by a single tester, a certified athletic trainer with 10 years of experience. Manufacturer and published guidelines for this instrument were used for participant positioning and device operation.^3,15 Previous studies indicated an intrarater reliability of 0.74 for talar inversion with this device. ³ Prior to data collection, the single rater’s reliability was established on the arthrometer using an intraclass correlation coefficient (ICC_2,1) and standard error of measurement (SEM). ³⁵ A total of 6 preliminary participants, not included in the study sample, were tested 3 times over 2 weeks on both limbs, and the ICC with SEM was calculated. The ICC_(2,1) for these participants was excellent (0.82-0.90), with an SEM of 2 to 4 degrees for inversion testing.

Participants were seated on the floor with the test leg extended and a bolster under the knee to approximate 15 degrees of flexion. The heel was placed in the device attachment, which was tightened to secure it firmly against the calcaneus with the ankle in approximately 10 degrees of plantarflexion. ³ The participants’ contralateral knee was flexed with the foot flat on the floor, and they were instructed to recline and relax their ankle musculature. The pressure actuator was positioned medially 5 cm proximal to the most prominent point on the medial malleolus, and a 150-N force was applied (Figure 1). Three trials were collected. Each trial was approximately 12 seconds in duration, with force applied for that length of time, per the manufacturer’s recommendations. A rest period of approximately 30 seconds was taken between trials as the device was reset. Participants were asked about pain during the trial, testing was stopped until pain resolved, and reminders, coupled with visual observation, were provided to participants to keep the ankle musculature relaxed. The force-displacement curve was visually inspected at the end of each trial for face validity. Both limbs were tested. The order of leg testing was randomized in the control and coper groups. The uninvolved or less symptomatic ankle was tested first in the chronic ankle instability group to minimize apprehension when testing the injured ankle. The single rater was not blinded to ankle injury history prior to testing but was blinded to CAIT scores and laxity and stiffness values.

OPEN IN VIEWER

Figure 1.

Depiction of inversion test setup with the instrumented arthrometer.

Data Reduction and Analysis

Utilizing the arthrometer’s software, maximum inversion was recorded, and the force-displacement curve data were extracted and calculated.^27,34 Stiffness was considered the relationship between a force applied and the displacement caused by that force. Linear stiffness was used as the most direct, accessible measure, given the information provided by the instrument and software. With linear stiffness, the force applied was proportional to the linear displacement. Calculating rotary stiffness may have been preferable, given the inversion rotational nature of the test, but assumptions required and possible errors resulting from estimating the torque made it undesirable. With rotational stiffness, the applied torque was proportional to the angular displacement of one end of the joint compared to the other. Due to the nonlinear tissue response, stiffness was calculated in the 40- to 60-N range and 125- to 150-N range using previous guidelines.^31,32 A spreadsheet document was exported via the arthrometer software, with columns reporting the applied force (N) and calculated inversion angle. Stiffness was determined as the slope of the line from 40 to 60 N and 125 to 150 N by calculating the change in the vertical axis divided by the change in the horizontal axis. Representative images for each group are displayed in Figure 2.

OPEN IN VIEWER

Figure 2.

Representative force-displacement curves for participants in (A) chronic ankle instability, (B) coper, and (C) control groups. Linear stiffness was determined as the slope of the line from 40 to 60 N and 125 to 150 N by calculating the change in the vertical axis divided by the change in the horizontal axis (force / strain). These are visual representations only. A spreadsheet document with numerical values was used to perform calculations.

Preliminary testing in our laboratory indicated that the first trial recorded by the arthrometer should be used as familiarization for each participant and was thus not included in the analysis. The mean of the second and third trials was calculated. Previous research has used the mean of 2 trials.^31,40 One-way analyses of variance (ANOVAs) were used to test for differences among the chronic ankle instability, coper, and control groups in demographics, laxity (inversion), and stiffness (α ≤ .05). When significant results were found, Tukey post hoc testing was applied to determine specific group differences (α ≤ .05). Confidence intervals and effect sizes were also calculated.

Results

Differences could not be detected between groups in age, height, or weight (Table 1). The chronic ankle instability group scored significantly lower on the CAIT than both the control and coper groups (P < .001). However, the control and coper groups did not have different CAIT scores. The chronic ankle instability group demonstrated significantly greater inversion than the control group (P = .05) (Table 2). With means reported, there were no differences in inversion between the chronic ankle instability and coper groups, or coper and control groups, and no significant differences between groups in stiffness at low-load (40-60 N) or high-load (125-150 N) ranges (P > .05) (Table 2).

OPEN IN VIEWER

Table 2.

Mean Values and 1-Way Analysis of Variance Results for Instrumented Arthrometer–Dependent Variables.

Variable	Mean ± SD	95% CI for Mean	P Value	Power (1 – β)	Effect Size (ηp2)	Effect Size for Multiple Comparisons (Cohen d)	95% CI for Effect Size (Cohen d)
Inversion talar tilt, deg
Chronic ankle instability	22 ± 12	17 to 28	.05	.57	0.12	Chronic ankle instability–coper: 0.49	−0.23 to 1.18
Coper	17 ± 10	12 to 22				Chronic ankle instability–control a : 0.87	0.12 to 1.57
Control	13 ± 9	8 to 18				Coper-control: 0.39	−0.32 to 1.07
Linear stiffness (40-60 N), N/mm
Chronic ankle instability	3 ± 1	3 to 4	.21	.32	0.07	Chronic ankle instability–coper: 0.59	−0.13 to 1.28
Coper	4 ± 1	4 to 4				Chronic ankle instability–control: 0.14	−0.56 to 0.83
Control	4 ± 1	3 to 4				Coper-control: 0.44	−0.27 to 1.13
Linear stiffness (125-150 N), N/mm
Chronic ankle instability	7 ± 1	6 to 7	.41	.20	0.04	Chronic ankle instability–coper: 0.43	−0.28 to 1.12
Coper	7 ± 1	7 to 8				Chronic ankle instability–control: 0.09	−0.60 to 0.78
Control	7 ± 1	6 to 7				Coper-control: 0.33	−0.38 to 1.01

Abbreviations: CI, confidence interval; SD, standard deviation.

a

Post hoc Tukey honestly significant difference indicates significant difference between groups at P < .05.

Discussion

The most important result was that the chronic ankle instability group demonstrated greater laxity than the control group on inversion with an instrumented arthrometer, as we hypothesized. However, the chronic ankle instability group did not display greater laxity than the coper group, which did not support our hypothesis. There was a high degree of variability in laxity among groups. Stiffness did not appear to be different between groups.

The chronic ankle instability group demonstrated increased laxity, or greater inversion, on the instrumented arthrometer compared to the control group but not the coper group. Based on these means, the control and coper groups were not statistically different. A systematic review concluded that unstable ankle groups demonstrated more inversion compared to healthy controls; however, the standardized effect sizes were often small and crossed zero, indicating limited differences. ¹ This supports our results in which only the chronic ankle instability–control comparison had a large effect size (0.87), while the chronic ankle instability–coper comparison had a moderate effect size (0.49), and the control-coper effect size was small (0.39). In the systematic review, the authors concluded that there was a lack of statistical differences between the chronic ankle instability and control groups, which does not support our results. We may have found differences by using stricter inclusion criteria than the studies included in the review. Using 2 more recent studies not included in the systematic review, we calculated the effect size and power for chronic ankle instability–control inversion laxity comparisons using tabled data. In both studies, the chronic ankle instability group demonstrated greater laxity than the control group, with effect sizes of 0.75 to 1.5 and powers of .53 to .99,^20,30 which more closely mirror our results. The difference in methods, including the type of arthrometer used,^{3,15,20,24,25,30} positioning of the foot/ankle, ²⁸ and inclusion/exclusion criteria, ² may have contributed to the inconsistency of results.

There were no group differences in linear stiffness in the 40- to 60-N low-load range, or the 125- to 150-N high-load range, during inversion. While the chronic ankle instability group had the lowest mean stiffness in both regions, the means and 95% confidence intervals reported were not statistically different from the other groups and were quite comparable to the control group. The coper group was the stiffest and had 95% confidence intervals that were shifted higher but still demonstrated overlap with the chronic ankle instability and control groups. Group comparisons indicated only small to moderate effect sizes, with all 95% confidence intervals for the effect size crossing zero. The low power may be attributable to the small sample size. With the means reported, our study appears to support no differences between groups in terms of stiffness, but there is limited literature on the stiffness of lateral ligaments in chronic ankle instability populations. ⁴⁰

One study reported no group differences between the chronic ankle instability, control, and coper groups in stiffness, ⁹ supporting our results. Decreased stiffness was found in cadaveric specimens after ligament sectioning ³² and in people with clinically evident mechanical instability compared to a mechanically stable group at 40 to 60 N and 200 N. ³¹ However, another study reported increased stiffness in the chronic ankle instability and coper groups to anterior drawer compared to uninjured controls. ⁴⁰

As a static joint stabilizer, a stiffer ligament could theoretically better respond to sudden inversion. Even after an injury and a substantial decrease in self-reported function, changes in stiffness were not detected with the means available. Our stiffness values were greater than those reported in a previous study that used the same units but a different arthrometer. ²⁹ Some authors calculated stiffness in the first 30% to 40% of the test,^31,32 or at different loads, ³¹ while another calculated it at the end range, ²⁸ and others did not specify.^9,29,40 Additionally, comparisons of stiffness using the anterior drawer test,^28,31,32 as opposed to inversion,^9,29 may not be appropriate. Finally, the rate of loading in a portable arthrometer does not approximate a potential real-world mechanism of injury. It is unclear if stiffness is a factor contributing to chronic ankle instability. Our results indicate that it is not. However, differences in acute treatment, immobilization, and rehabilitation after injuries could affect stiffness values. Prospective research following initial injuries is necessary to determine the clinical course of chronic ankle instability.

Only a few studies to date have incorporated a coper group into comparisons of laxity, but they may better represent a clinically useful pathway from injury to healing.^38,39 Separating out copers from controls may improve the effect size of group comparisons. Our results indicate that copers appear to exist on a continuum, somewhere between those with chronic ankle instability and controls. Using previous studies’ tabled data, chronic ankle instability groups were more lax to inversion rotation than coper groups, with large effect sizes (1.03-1.4) and powers (.81-.96) in studies utilizing a different arthrometer.^19,30 With the means reported, we did not find significant differences between the chronic ankle instability and coper groups in laxity and stiffness, but they had moderate effect sizes, indicating possible clinical relevance. ¹⁶ Based on our means, it appears that the coper group falls in between the chronic ankle instability and control groups in terms of laxity but may skew stiffer. Copers’ mean stiffness was higher at low- and high-load regions of the force-displacement curve but not significantly so (P > .05).

Comparisons between coper and control groups are also mixed. Our coper group’s mean inversion value was larger than that of the control group, but the means were not statistically different, with only a small to moderate effect size (0.39). Utilizing tabled data, a control-coper comparison of mean inversion demonstrated a small effect size (0.16) and low power (.07) in one study, ³⁰ while another study reported no control-coper group differences. ⁹ Both of these studies support our results. Our small effect size and overlap in 95% confidence intervals do not support differences in stiffness between copers and controls. Little is known about copers following an injury. They may demonstrate better tissue response, especially at low loads where others have found differences ³¹ in being able to respond appropriately and moderate inversion forces.

Currently, it appears that copers overlap substantially in terms of laxity and stiffness with both the chronic ankle instability and control groups. Thus, mechanical laxity and stiffness do not appear to be contributing factors driving individuals toward or away from chronic ankle instability after an initial sprain. Copers reported better function than those in the chronic ankle instability group and similar function to controls but had a wide range of laxity and stiffness values, some of which were comparable to participants with chronic ankle instability. These results also provide evidence for a model of chronic ankle instability with more subgroups ¹² in which mechanical laxity is separated out from perceived instability and recurrent sprains while also addressing the combined effects of those characteristics.

There are several limitations to this study. The first is that it was cross-sectional; thus, we were unable to determine if the laxity and stiffness values noted are results of ankle sprains and chronic ankle instability or if pre-existing laxity and stiffness caused the participants to develop chronic ankle instability or be copers. There was no specific “numerical” indicator of instability, as we utilized a self-report questionnaire and an instrumented arthrometer to perform talocrural inversion rather than radiological measures. The talocrural inversion was used because it is the recommended test with the arthrometer and software rather than the anterior drawer test. Thus, rotational instability was tested, and some instability may not be picked up with any method of testing, as it does not replicate sporting or daily living conditions. Additionally, participant differences in neuromuscular control, severity of the initial injury, rehabilitation, and involvement of the anterior talofibular versus calcaneofibular ligament were not determined but could have affected results. We did not use electromyography and could not tell if the ankle musculature was truly relaxed during arthrometry, which may have influenced laxity and stiffness values. Preliminary testing and previous studies^31,40 indicated that the first data collection trial was not reliable and was lower than subsequent trials. We discarded this trial, but its lack of inclusion could have affected our results. We calculated linear stiffness based on available data, but rotary stiffness may present different results. An a priori power analysis determined the sample size, but the stiffness measures were underpowered. A larger sample size would likely increase power.

Conclusion

The chronic ankle instability group demonstrated greater inversion mechanical laxity to the instrumented arthrometer compared to the control group as well as decreased self-reported function compared to the control and coper groups. However, the coper group was not more lax than the chronic ankle instability or control groups. Increased mechanical laxity appears to exist in the chronic ankle instability group compared to the control group. Copers’ laxity appears to exist on a continuum between the chronic ankle instability and control groups. There were no group differences in linear stiffness, and tissue response to the applied load may not be a factor contributing to chronic ankle instability. Future prospective studies should determine the role of mechanical laxity in developing and perpetuating chronic ankle instability.