Very few studies exist on the topic of balance and offloading orthotic devices, despite the wide use of offloading devices in orthopedic and neurologic populations. For example, current treatment of the diabetic foot ulcer is to offload the affected limb using ankle-foot orthoses, such as the Total Contact Cast, felt and foam, a wedged shoe, walking boot (WB; also referred to as a removable cast walker), or the instant Total Contact Cast.1,2 The WB is a removable instant total contact cast,3 which mimics the Total Contact Cast (considered the criterion for wound healing)1,2 by redistributing pressure over the entire lower leg and foot. These devices avoid point pressures4 and limit tibial progression during gait to avoid forefoot loading.5
A WB orthotic device may contribute to impaired postural control.6 Impaired postural control with a WB may exist because it produces an increased leg length discrepancy7 from the thick rocker bottom sole, which also effectively reduces the base of support.6,8 Moreover, the WB limits ankle range of motion, which decreases proprioceptive feedback from ankle joint receptors and muscle spindles spanning the ankle joint and decreases the ability to use an ankle strategy in balance corrections on the involved limb.9,10
Although the few existing studies on this topic provide valuable information, there are still many unanswered questions that warrant further investigation. For example, Lavery et al.11 show that in patients with diabetes, quiet stance center of pressure travel was not significantly different across conditions of donning standard shoes, a removable cast walker, a half-wedged bottom shoe, or a total contact cast. In contrast, a different study found that a rocker bottom alone negatively influenced balance responses in healthy individuals during perturbed stance (i.e., increased sway amplitude, sway variance, time to peak sway).8 Moreover, a recent study pointed to the need for more research because the authors found a trend that center of mass (CoM) motion increased with removable cast walker strut height and suggested there may be stabilizing effects of an ankle-high WB, but these results specific to postural control did not reach statistical significance.12 Thus, it is unknown if a WB enhances, diminishes, or has a neutral effect on balance. Furthermore, there is a general need for more understanding of how offloading devices affect balance6 and a need to measure balance in more challenging conditions, as suggested by the previous studies on this topic.8,11,12
Therefore, the primary hypothesis tested in the current study was that a WB increases body motion across a range of simple to challenging conditions. The secondary hypothesis was that adding a heel lift to the noninvolved limb would reduce body motion by correcting the leg length discrepancy. Testing these hypotheses address clinically relevant questions and fill a gap in our knowledge about the impact of a WB on balance control.
Twelve healthy young subjects (mean age, 27 ± 5.06 years; 6 males, 6 females; height, 171.32 ± 11.05 cm; weight, 70.4 ± 13.9 kg) were recruited to participate in this study. All subjects were free from balance disorders, did not have a leg length discrepancy of >2.5 cm, and had not worn a WB in the previous 10 years. Subjects gave written informed consent, and all tests were approved by the Institutional Review Board at the University of Hartford.
In each test, subjects wore comfortable athletic shoes (i.e., control), a WB, or a WB and a heel lift in the contralateral shoe. The WB was nonpneumatic, with a fixed ankle (09910, Truelife Prosthetic & Orthotics, Dublin Ireland). The WB extended up most of the lower leg but allowed free knee movement. One of three standard sizes was selected for each subject, such that the strut length relative to leg length was approximately uniform across subjects (i.e., proximal to the apex of the calf muscle). Subjects performed three types of balance tests designed to range in difficulty and to assess various aspects of balance control (Figure 1). In each test, upper-body (UB) and lower-body motion in the sagittal and frontal plane was measured using accelerometer-based dual-axis tilt sensors (Crossbow, CA). One sensor was attached between the two scapulae and the other at the inferior border of subjects’ greater trochanter, identical to a previous study.13 Data collection was 100 samples/s.
Within each type of balance test (i.e., quiet stance, functional reach, or walking), subjects were randomly assigned an orthotic condition, athletic shoes, WB, or WB plus heel lift. Then, the subject randomly completed all tests with that particular orthotic condition. The subject was then randomly assigned to a different orthotic condition and completed, in random order, all tests. The WB was always worn on the right leg. For the heel lift conditions, one or more precut foam pads were inserted into the shoe to level the pelvis in the frontal plane. A digital posture alignment measurement system (Yuki Trading, Tokyo, Japan) was used to quantify the pelvic angle in control (0.66 ± 1.4 degrees) with the WB alone (–1.6 ± 2.5 SD, where negative indicates the WB side was higher) and with the addition of a heel lift in the contralateral shoe (0.42 ± 1.4 degrees).
BALANCE TEST 1: QUIET STANCE TESTS
The first balance test was quiet stance in four conditions: 1) eyes open on a firm surface; 2) eyes closed on a firm surface; 3) eyes open on foam surface; and 4) eyes closed on foam surface. This test was selected to investigate sensorimotor integration processes for balance,14 because the different surfaces and visual availability require subjects to shift reliance to different sensory systems. Each quiet stance test was 50 s. In quiet stance tests, subjects crossed their arms over their chest in a comfortable position and wore headphones that played a story to eliminate any background noise and to encourage similar cognitive processes throughout each test. Subjects were instructed to stand upright and natural. For quiet stance tests, we used outputs from the sensors to assess balance. Two commonly used balance measurements were calculated: CoM root-mean-square (RMS) tilt with respect to vertical and tilt velocity.15 Larger RMS values represent greater body motion. To provide a simple single dependent output variable, we summed RMS values between anterior-posterior (AP) and medial-lateral (ML) motion. CoM was estimated from UB and lower-body motion using a four-link model for frontal plane motion13,16 and a two-link model for sagittal plane motion17 in conjunction with anthropometric measures.18
BALANCE TEST 2: FUNCTIONAL REACH TESTS
The second balance test was a functional reach test.19 This test was performed in three directions (anterior, right, and left) and was selected to investigate the extent to which subjects could compensate for limited ankle range of motion with the WB when voluntarily extending to their limits of stability. Subjects were instructed to reach as far as possible in a specific direction without lifting their feet off the ground three times. The largest value reached was recorded with a ruler.
BALANCE TEST 3: PERTURBED AND UNPERTURBED WALKING TESTS
The third balance test was a walking test in perturbed and unperturbed conditions for 90 s. In all walking tests, subjects crossed their arms over their chest in a comfortable position and wore headphones that played a story. Subjects were instructed to walk naturally and respond naturally to perturbations when they were delivered. We selected a treadmill speed of 3.3 km/hr for all walking tests. All subjects reported that 3.3 km/hr was comfortable, even though some would have self-selected a different speed. Before testing with the WB, subjects were provided time to walk in the WB.
The perturbed walking test was a novel experimental paradigm selected to investigate balance in more dynamic and challenging contexts, as suggested by others.8,11,12 Certain balance consequences may only be detectable in more challenging conditions, consistent with previous studies using continuous external perturbations to assess balance during treadmill walking.20,21 The perturbed tests also provided greater detail into balance because responses to perturbations at specific frequencies could be quantified. In perturbed walking tests, the platform rotated continuously, which was perceived by subjects as walking on a destabilizing and unpredictable surface. Subjects reported that their balance was challenged but also that the test was not too difficult or scary. During these tests, the platform rotated according to an in-phase sum-of-sines waveform (0.5 and 1.5 Hz), with a peak amplitude of 1.5 degrees. This platform rotation resulted in primarily frontal plane perturbations [sin(θ) ∼ θ for θ < 10 degrees]. The peak surface translation delivered to subjects was ∼2.2 cm in the frontal plane and only ∼0.029 cm in the sagittal plane. Previous studies indicate that frontal plane stability involves more active neural control than sagittal plane20 and frontal plane motion during gait may be particularly sensitive to orthotic interventions12, 22 and external perturbations.20,21 Each subject’s position on the treadmill was 85 cm away from the platform center. Platform velocity was measured using a high-precision gyroscope (Watson Industries, WI).
For all walking tests, the dependent variables were UB motion in the sagittal and frontal planes. To gain more information of balance responses in perturbed walking tests, discrete Fourier Transforms were used to characterize UB tilt in the frequency domain. A frequency domain analysis decomposes data in the time domain into its frequency components. This approach enables us to tease apart meaningful rhythmic behavior that only occur at specific frequencies. In the current study, we calculated how the orthotic condition influenced the UB cadence. UB cadence was defined as the frequency at which UB ML motion had the largest amplitude and corresponds to the frequency of natural oscillatory UB motion during walking. We also calculated how the orthotic condition influenced the distribution of UB motion across frequencies. UB motion present across a wide range of frequencies was considered less rhythmic than motion at a single frequency because signal processing theory defines white noise as spectral power distributed equally across a specific wide bandwidth of frequencies.23 Therefore, the percent of nonrhythmic motion was defined as the percentage of UB ML motion amplitude outside of the UB cadence frequency (calculated across 0.014–to 5.4 Hz).
For quiet stance, a separate repeated-measures analysis of variance was performed for two variables: RMS CoM tilt and tilt velocity (summed across frontal and sagittal planes). Model effects were vision (eyes open, closed), surface (foam, firm), and orthotic condition (WB, WB plus heel lift, control). For functional reach, a separate repeated-measures analysis of variance was performed for anterior and lateral directions. In anterior tests, the orthotic condition was the model effect; in lateral tests, the orthotic condition and direction of reach were model effects. For walking tests, a separate repeated-measures analysis of variance was performed for four variables: UB ML RMS motion, UB AP RMS motion, UB cadence, and the percent of nonrhythmic motion. In UB cadence and RMS variables, model effects were platform motion (perturbing, nonperturbing) and orthotic condition. In the percent of nonrhythmic motion variable, the model effect was orthotic condition. Multiple comparisons were made with Tukey’s post hoc analyses. JMP statistical software was used. In all statistical tests, the significance level was set to p < 0.05.
Quiet stance results are summarized in Figure 2 for CoM tilt. Across surface types and visual conditions, tilt generally increased with the WB compared with control and decreased with the heel lift compared with the WB. Surface type (p < 0.001), visual availability (p < 0.001), and orthotic condition (p < 0.001) had significant effects on RMS tilt and tilt velocity. RMS tilt was significantly larger with the WB compared with control. With the heel lift, tilt was significantly lower than the WB alone and not significantly different from the control. There was a significant interaction effect (p = 0.037) between surface and orthotic condition, whereby the orthotic condition had a significantly larger effect on tilt on the foam compared with firm surface. Similarly, there was a significant interaction effect (p = 0.037) between visual availability and orthotic condition, whereby the orthotic condition had a significantly larger effect on tilt in eyes closed compared with open. Thus, in conditions that are associated with shifting reliance to vertical orienting sensory cues, the WB had the largest effect on subjects’ tilt.
For RMS tilt velocity, both WB and the heel lift conditions resulted in significantly greater velocity than control. There was a significant interaction effect (p = 0.031) between surface and orthotic condition, whereby the orthotic condition had a significantly larger effect on tilt velocity on the foam compared with the firm surface.
Results from functional reach tests are summarized in Figure 3. There was a significant effect (p = 0.018) of the orthotic condition in anterior reach. In both orthotic condition and heel lift condition, anterior reach was significantly smaller compared with control. In the lateral direction, there was a small but significantly larger (p = 0.037) lateral reach to the right compared with the left (1.7 cm larger). However, orthotic condition did not have a significant effect on lateral reach. Thus, during the functional reach test, subjects were able to compensate for the WB during the lateral reach but not during the anterior reach.
PERTURBED AND UNPERTURBED WALKING TESTS
Figure 4A shows an example of platform motion and a representative subject’s walking data for one nonperturbed and one perturbed control condition. This figure illustrates that the continuous platform movement effectively perturbed balance evoking larger UB motion in the perturbed compared with unperturbed conditions. Across subjects, there was significantly larger UB AP (p = 0.0013) and ML (p < 0.001) motion (Figure 4 B). RMS UB ML and AP motion with the WB on was significantly larger compared with control. UB motion with the heel lift was on average lower compared with the WB but was not significantly different from either WB or control conditions in the AP direction and was significantly larger compared with control in the ML direction.
During perturbed walking, subjects’ UB motion included prominent peaks in amplitude at 0.5 and 1.5 Hz, corresponding to the platform movement, and a peak around 0.8 Hz corresponding to the subject’s UB cadence. In unperturbed walking, only one prominent peak was present at the UB cadence.
Perturbed walking significantly increased UB cadence (p < 0.001). Orthotic condition did not have a significant effect on UB cadence, but on average, the shift in UB cadence was larger with the WB compared with the control and heel lift (Figure 5). In unperturbed (but not perturbed) walking, UB motion was significantly less rhythmic in the WB and heel lift conditions compared with control (p < 0.001).
WB INFLUENCE ON BALANCE
Results from the current study support the hypothesis that a WB orthotic offloading device increases body motion across a range of simple to difficult balance tests. Possible reasons include the following: 1) reduced surface contact with the floor via the rocker bottom; 2) increased leg length discrepancy; and 3) decreased ankle joint motion, which results in less proprioceptive feedback around the ankle, less ability to generate an ankle strategy for balance, and less tibial progression during gait. In each balance test, one or more of these factors likely had an impact on the measures of balance, as described below.
During quiet stance, the rocker bottom may have contributed to increased body motion because it reduces surface contact area and, consequently, reduces the effective base of support,6,8 which is known to alter postural responses.9 Body motion was also influenced by the leg length discrepancy caused by the WB because the heel lift significantly decreased tilt, similar to previous studies.7 Finally, the significant interaction effects suggest that the WB negatively influenced balance most in quiet stance conditions that required subjects to shift reliance toward vestibular or visual feedback, or both14 (i.e., greater increases in foam compared with firm surfaces and in eyes closed compared with open). From a theoretical perspective, the WB produces a passively stiff ankle joint.6 This passive stiffness tends to orient the lower body perpendicular to the surface and away from upright when the surface is not horizontal,16,17 such as foam. Similar to foam, the rocker bottom produces a surface that is not horizontal. Thus, the passive ankle stiffness associated with the WB could have oriented the body away from vertical and contributed to larger quiet stance motion, which were not fully compensated for with greater contributions from vision or vestibular.16,17
In functional reach tests, the heel lift did not significantly increase reach compared with WB alone, indicating that a leg length discrepancy probably did not contribute to a reduced functional reach. Therefore, the lower anterior reaches with the WB were probably influenced by the reduction in ankle range of motion and rocker bottom. The rocker bottom may have felt like an unstable support when reaching as far as possible. A WB is specifically designed to reduce ankle range of motion. Thus, the only way subjects would not have a reduced functional reach with WB would be to increase motion about other joints of the body. Our results indicate that subjects did not fully compensate. The functional reach test has predictive validity for fall risk in older adults, such that reduced reach is associated with greater fall risk.24 If future studies find that older adults have reductions in reach with the WB, then older adults wearing the WB may have a heightened fall risk.
In the walking tests, several characteristics of the WB may explain the differences in UB motion (Figure 4). The WB was designed to limit ankle joint motion and tibial progression only in the leg with the WB. This asymmetrical biomechanical change likely contributed to an asymmetrical gait pattern that may have increased UB motion. Moreover, the restricted ankle joint motion interfered with subjects’ ability to make ankle strategy balance corrections during the perturbed tests, consistent with previous studies that identified a lateral ankle strategy as one of two methods to compensate for lateral perturbations during gait.10 Across orthotic conditions, balance responses coincided with the stimulus frequencies, similar to standing balance studies,16,17 which indicates that the WB did not interfere with this balance strategy. The shift toward higher-UB cadence in the perturbed walking condition was also similar across orthotic conditions. UB cadence is directly related to gait cadence, which indicates that subjects adopted a strategy of taking shorter steps when perturbed, in agreement with previous studies in healthy and diabetic patients.21,25 Although the current study demonstrated an effective method to continuously perturb balance during gait, perturbations at other stimulus frequencies or at specific phases in the gait cycle26 could be evaluated.
CLINICAL RELEVANCE AND HEEL LIFT
The offloading device tested in the current study is used especially by patients with orthopedic complications and diabetic foot ulcers because of the established effectiveness in healing.1,2 If future studies with patients continue to show increased body motion with offloading devices, then patients should be educated on fall risk management and hospitals/clinics should factor in the presence of offloading devices when assessing fall risk. This would be especially important for the diabetic population, which affects >250 million people worldwide and commonly results in multiple health complications, including sensory neuropathy (loss of protective sensation), which impairs balance.27,28 Thus, future studies should include subjects with diabetic peripheral neuropathy,29 who are often prescribed an offloading device.
One solution to partially restore balance was investigated in the current study. We lessened the leg length discrepancy by incorporating a heel lift, which resulted in either similar or reduced tilt compared with the WB alone. In general, correcting a leg length discrepancy could enhance comfort and reduce the risk of long-term orthopedic complications.30,31 A heel lift, as opposed to a full foot lift, was used because leg length discrepancies were <2.5 cm, consistent with many clinical practice methods. However, most subjects reported discomfort with the heel lifts because the heel lifts were not skived to reduce discontinuities between the heel lift and insole. Consistent with this notion, the heel lift was most effective in quiet stance, in which less overall movement and weight shifting was required compared with functional reach or walking. Thus, a more comfortable heel lift may provide improved results. An alternative explanation for the diminished effectiveness of the heel lift in functional reach and walking may be that other WB characteristics (e.g., the reduced ankle motion or rocker bottom) were more detrimental in these conditions.
1. Armstrong DG, Boutlon AJM, Lavery LA, et al. Evaluation of removable and irremovable cast walkers in the healing of diabetic foot wounds: a randomized controlled trial. Diabetes Care
2005; 28: 551–558.
2. Piaggesi A, Macchiarini S, Rizzo L, et al. An off-the-shelf instant contact casting device for the management of diabetic foot ulcers. Diabetes Care
2007; 30: 586–590.
3. Armstrong DG, Short B, Nixon BP, et al. Technique for fabrication of an “instant” total contact cast for treatment of neuropathic diabetic foot ulcers. J Am Podiatr Med Assoc
2002; 92: 405–408.
4. Fleischli JG, Lavery LA, Vela SA, et al. Comparison of strategies for reducing pressure at the site of neuropathic ulcers. J Am Podiatr Med Assoc
1997; 87: 466–472.
5. Lavery LA. 5 Questions and answers about off-loading. Adv Skin Wound Care
2003; 16: 231–234.
6. Van Deursen R. Footwear for the neuropathic patient: off-loading and stability. Diabetes Metab Res Rev
2008; 24: 96–100.
7. Mahar RK, Kirby RL, MacLeod DA. Simulated leg-length discrepancy: its effect on mean center-of-pressure position and postural sway. Arch Phys Med Rehabil
1985; 66: 822–824.
8. Albright BC, Woodhull-Smith WM. Rocker bottom soles alter the postural response to backward translation during stance. Gait Posture
2009; 30: 45–49.
9. Horak FB, Nashner LM. Central programming of postural movements: adaptation to altered support-surface configurations. J Neurophysiol
1986; 55: 1369–1381.
10. Hof AL, Vermerris SM, Gjaltema WA. Balance responses to lateral perturbations in human treadmill walking. J Exp Biol
2010; 213: 2655–2664.
11. Lavery L, Fleishli JG, Laughlin TJ, et al. Is postural instability exacerbated by off-loading devices in high risk diabetics with foot ulcers. Ostomy/Wound Manage
1998; 34: 26–34.
12. Crews RT, Saveed F, Najafi B. Impact of strut height on offloading capacity of removable cast walkers. Clin Biomech
2012; 7: 725–730.
13. Goodworth AD, Chandan A, Chase H, et al. Stance width influences frontal plane balance responses to centripetal accelerations. Gait Posture
2013; 37: 98–102.
14. Cohen H, Blatchly CA, Gombash LL. A study of the clinical test of sensory interaction and balance. Phys Ther
1993; 73: 346–354.
15. Maurer C, Peterka RJ. A new interpretation of spontaneous sway measures based on a simple model of human postural control. J Neurophysiol
2005; 93: 189–200.
16. Goodworth AD, Peterka RJ. Sensorimotor integration for multi-segmental frontal plane balance control in humans. J Neurophysiol
2012; 107: 12–28.
17. Peterka RJ. Sensorimotor integration in human postural control. J Neurophysiol
2002; 88: 1097–118.
18. Winter DA. Biomechanics and Motor Control of Human Movement
. Hoboken, NJ: Wiley; 2005.
19. Duncan PW, Weiner DK, Chandler J, Studenski S. Functional Reach: a new clinical measure of balance. J Gerontol
1990; 45: M192–M197.
20. O’Connor SM, Kuo AD. Direction-dependent control of balance during walking and standing. J Neurophysiol
2009; 102: 1411–1419.
21. McAndrew PM, Dingwell JB, Wilken JM. Walking variability during continuous pseudo-random oscillations of the support surface and visual field. J Biomech
2010; 43: 1470–1475.
22. Najafi B, Miller D, Jarrett BD, et al. Does footwear type impact the number of steps required to reach gait steady state?: an innovative look at the impact of foot orthoses on gait initiation. Gait Posture
2010; 32: 29–33.
23. Bendat JS, Piersol AG. Random Data: Analysis and Measurement Procedures
. New York: Wiley; 2000.
24. Duncan PW, Studenski S, Chandler J, et al. Functional reach: predictive validity in a sample of elderly male veterans. J Gerontol
. 1992; 47: M93–M98.
25. Menz HB, Lord SR, St George R, et al. Walking stability and sensorimotor function in older people with diabetic peripheral neuropathy. Arch Phys Med Rehabil
2004; 84: 245–252.
26. Oddsson LIE, Wall C III, McPartland MD, et al. Recovery from perturbations during paced walking. Gait Posture
2004; 19: 24–34.
27. Bonnet C, Carello C, Turvey MT. Diabetes and postural stability: review and hypotheses. J Mot Behav
2009; 41: 172–187.
28. Bloomgarden ZT. Diabetes complications. Diabetes Care
2004; 27: 1506–1514.
29. Higmans JM, Geertzen JHB, Dijkstra PU, et al. A systematic review of the effects of shoes and other ankle or foot appliances on balance in older people and people with peripheral nervous system disorders. Gait Posture
2007; 25: 316–323.
30. Caselli M. Evaluation and management of leg-length discrepancy. Podiatry Manage
2006; 25: 147–156.
31. Gurney B. Leg length discrepancy. Gait Posture
2002; 15: 195–206.