Secondary Logo

Journal Logo

Original Research Articles

Spinal and Pelvic Kinematics During Gait in People with Lower-Limb Amputation, with and without Low Back Pain: An Exploratory Study

Devan, Hemakumar PhD; Dillon, Michael P. PhD; Carman, Allan B. PhD; Hendrick, Paul PhD; Hale, Leigh PhD; Boocock, Mark PhD; Ribeiro, Daniel Cury PhD

Author Information
Journal of Prosthetics and Orthotics: July 2017 - Volume 29 - Issue 3 - p 121-129
doi: 10.1097/JPO.0000000000000137
  • Free


Low back pain (LBP) is a common musculoskeletal impairment after transfemoral amputation (TFA) and transtibial amputation (TTA).1,2 Low back pain affects between 55% and 71% of people with TFA and TTA,1 twice the estimated LBP prevalence (31%) reported in the population of persons without amputation.3 Although LBP is often thought to be more common in people with TFA,4 studies including people with TTA report a similar prevalence.5,6 Low back pain has also been shown to restrict daily activities5 and is associated with poor physical quality of life7 in people with lower-limb amputation.

Low back pain is a multifactorial impairment with physical (biomechanical), psychosocial, and personal risk factors.8 Among physical factors, spinal and pelvic movement patterns in people with lower-limb amputation such as increased trunk lateral flexion toward the prosthetic limb and increased anterior pelvic tilt are thought to be a major contributor to LBP.9,10 Such movement adaptations to the patterns of trunk lateral flexion and trunk postures are well-established risk factors for LBP in the population of persons without amputation.11 Some of these spinal movement patterns could be classified as “maladaptive” and could lead to long-term changes in passive ligamentous structures of the lumbopelvic region resulting in fatigue of spinal musculature and, theoretically, increased risk of LBP in people with lower-limb amputation.12

Few studies have investigated three-dimensional (3D) spinal movement patterns in people with TFA and LBP.13,14 None have included people with TTA. In discussing their results, authors have not clearly described the link between movement patterns observed and the presence of LBP. For example, a recent study of people with TFA14 found differences in lumbar movement patterns in the sagittal and transverse plane between the LBP and No-LBP groups14 but did not explain how these differences potentially contributed to the onset and maintenance of LBP.

Given the high prevalence of LBP in people with TFA and TTA1,5 and limited previous research investigating spinal motion in individuals with lower-limb amputation and LBP, there is a need to describe the kinematics of thoracic, lumbar, and pelvic segments in persons with TFA and TTA, with and without LBP, and in doing so, develop hypotheses that explain how the movement patterns may affect the onset and maintenance of LBP so as to guide future studies of cause and effect.



This project was approved by the University of Otago Ethics Committee and all participants provided informed consent in writing.

The National Artificial Limb Service (NZALS), the entity that provides funding for prosthetic services, invited people that met the a priori inclusion criteria to participate in the study. A letter of invitation was sent to all the potential participants who met the inclusion criteria.

Participants were included if they were between 18 and 65 years of age, with unilateral TFA or TTA due to trauma, congenital causes, or tumors. Participants were excluded if they used walking aids during the last year. Participants were also excluded if they had spinal surgery in the last 12 months and/or currently under medications for inflammatory conditions (e.g., rheumatoid arthritis and ankylosing spondylitis). Participants with dysvascular amputation were not included. People with dysvascular amputation were excluded because of the confounding influence systematic illnesses (e.g., cardiovascular problems, deficits in balance, and muscle strength) on measures of LBP.


The principal investigator (H.D.) collected the data at research centers in Dunedin and Auckland. In Dunedin, kinematic data were collected using a 12-camera motion analysis system (Motion Analysis Corporation TM, Santa Rosa, CA, USA). In Auckland, a 9-camera system (Qualisys Medical AB, Sweden) was used. Both research centers were equipped with two AMTI (AMTI, Inc, Watertown, MA, USA) force plates embedded in series in the center of a 7-m walkway. Kinematic and external force data were collected at 120 Hz and 2400 Hz, respectively.


Before gait analysis, participants completed a series of standardized evaluations of LBP prevalence, frequency, site, and bothersomeness, which were recommended for LBP prevalence studies.15 Specifically, LBP disability was measured using the 24-item Roland Morris Back Pain Questionnaire.16 Prosthetic mobility was recorded using a 14-item Locomotor Capability Index (LCI),17 which measured basic and advanced prosthetic mobility.17 The socket comfort was quantified using the Socket Comfort Score.18 The presence of anxiety and depression were assessed using the Hospital Anxiety and Depression Scale.19

Participants were classified into a “LBP” group if they answered “yes” to the question, “Have you had persistent bothering LBP for the past 12 months?”15 A time frame of 12 months was chosen owing to the chronic, intermittent nature of LBP in this population.2,5

Anthropometric measures were recorded including height, weight, and leg length. Leg length was measured bilaterally as the distance between greater trochanter and lateral malleolus of ankle and a corresponding landmark on the prosthetic foot in standing.

Fifty-four retro-reflective markers (12.7 mm in diameter) were placed over standardized anatomical landmarks (or corresponding landmarks on the prosthesis) in accordance with the model described by Intolo et al.20 (Figure 1). All markers were placed by the same investigator (H.D.) to minimize marker displacement error.

Figure 1:
Marker set used for the study (n = 54). Trunk segment: Xiphoid process (PX), suprasternal notch (IJ), right and left acromion process (R.SHO, L.SHO), spinous processes of T1, T6, T12, and L3, right and left inferior angle of scapula (R.INF, L.INF), 3 cm to the right (R.L3) and left of L3 spinous process (L.L3), midpoint of PX and T12 (MID-PX-T12). Pelvic segment: Iliac crest (R.ILI, L.ILI), anterior and posterior superior iliac spine (R.ASI, L.ASI). Thigh segment: Greater trochanter (R.GT, L.GT), mid-anterior thigh (RT1, LT1), lateral one third (RT2, LT2), lateral two thirds (RT3, LT3), posterior thigh (RT4, LT4), medial and lateral epicondyle (RKNE, LKNE). Shank segment: Mid-anterior shank (RS1, LS1), lateral one third (RS2, LS2), lateral two thirds (RS3, LS3), posterior shank (RS4, LS4), medial and lateral malleolus (RANK, LANK). Foot segment: Base of first (RFT1, LFT1) and fifth metatarsals (RFT2, LFT2) and calcaneus (RFT3, LFT3).

A static trial was recorded with the subject in standing position. The medial knee and ankle markers were removed before walking trials. Participants walked at their self-selected walking speed. Three practice trials were given to familiarize participants with the walking task. Data were subsequently collected until a minimum of five successful trials; that is, when two subsequent prosthetic heel strikes occurred over the force plates. The gait cycle was defined as prosthetic heel contact to ipsilateral prosthetic heel contact.


Raw kinematic data were low-pass filtered using a fourth-order Butterworth digital filter at 6-Hz cut off. A 3D link-segment kinematic model was constructed in Visual 3D (C Motion, Inc, USA). The kinematic model comprised 10 segments including thoracic, lumbar, trunk, pelvis, bilateral thigh, shank, and foot segments. Definitions of local coordinate systems were in keeping with the recommendations of the International Society of Biomechanics (ISB),21 except for the pelvis and trunk. For the pelvic segment, a Visual 3D approach22 was used where the iliac crest markers defined the proximal end of the segment and the greater trochanter markers defined the distal end of the segment (Figure 1). These landmarks were chosen for pragmatic reasons as they are easy to locate by palpation. Furthermore, the relative vertical orientation of iliac crest and greater trochanter marker provides a direct measurement of pelvic tilt in the sagittal plane. This approach is a commonly used method for clinical measurement of pelvic tilt and thereby suggested to improve repeatability by the developers of Visual 3D.22 For the trunk segment, the acromion process markers defined the proximal end of the segment and the iliac crest defined the distal end of the segment.23 Tracking markers for the trunk segment included both thoracic (T1, T6, T12, PX, IJ, R.INF, and L.INF) and lumbar markers (L3, R.L3, and L.L3) (Figure 1).


Thoracic and lumbar segment angles were described with respect to the segment immediately inferior.20 Motion of the pelvic segment was described with reference to the laboratory global coordinate system. The trunk segment was described with reference to the pelvic segment.24 All the joint angles were expressed relative to the static trial and presented as angle-time series curves normalized to 100% of the prosthetic limb gait cycle. A Cardan rotational sequence (Z-X-Y) was used in computing kinematics for all the joint segments where Z-X-Y define motion about the sagittal-coronal-transverse planes respectively.


Because of the small sample size, we adopted a step-by-step approach before calculating the ensemble averages of angle-time curves of thorax, lumbar, pelvis, and hip joint segments. First, individual trials were inspected within each participant for consistency. Erroneous trials were removed based on visual inspection before averaging trials within each participant. Next, the mean curves of individual participants in “LBP” and “No-LBP” groups were plotted against each other for further verification. Then, the average trial for each participant was ensemble into an LBP and No-LBP group for those with TFA and TTA. As a final step, we compared the mean curves of from our study with similar previous studies13,25–28 to ensure that the range and pattern of movement curves are consistent. This step-by-step approach ensured that the mean curves of thorax, lumbar, and pelvic segments of each group actually represent the movement patterns of individual participants within the group. Joint/segment angles in all three planes were presented as a mean and one standard deviation.


Participant Characteristics

A total of 18 participants with lower-limb amputation (6 TFA, 13 TTA) were recruited (Table 1). The cohort was predominantly male and relatively young with high mobility and socket comfort (Table 1). None of the participants reported Hospital Anxiety and Depression scores indicative of depression (Table 1).

Table 1:
Participant characteristics

In comparison to TFA-No LBP group, persons with TFA-LBP were middle-aged men, with many more years of prosthesis use and walked at a slower velocity (Table 1). By comparison, the TTA-LBP and TTA-No LBP groups were very similar (Table 1).

Both TFA-LBP and TTA-LBP participants presented with mild LBP intensity, and low levels of LBP disability as indicated by Roland Morris Disability Scores29 (Table 1).

Persons with TFA

There were differences observed in spinal (thoracic and lumbar) and pelvic movement patterns between TFA-LBP and TFA-No LBP groups (Figures 2–4).

Figure 2:
Coronal plane kinematics in persons with TFA.
Figure 3:
Transverse plane kinematics in persons with TFA.
Figure 4:
Sagittal plane kinematics in persons with TFA.

In the coronal plane (Figure 2), both TFA-LBP and TFA-No LBP groups exhibited similar trunk lateral flexion toward the prosthetic limb during prosthetic stance (Figure 2C) but used different strategies to position the trunk. In the TFA-LBP group, people maintained a neutral lumbar spine (Figure 2B), and elevated the pelvis on the intact side (Figure 2D) to effectively “tip” the trunk laterally toward the prosthetic limb. By contrast, the TFA-No LBP group maintained a more horizontal pelvis (Figure 2D), and laterally flexed the lumbar spine toward the prosthetic limb (Figure 2B) to achieve the same trunk position.

In the transverse plane (Figure 3), the TFA-LBP group showed increased lumbar rotation toward the prosthetic limb that persisted through the gait cycle (Figure 3B); that is, the lumbar segment did not rotate toward the intact limb during the gait cycle as it did in the TFA-No LBP group (Figure 3B).

In the sagittal plane (Figure 4), the TFA-LBP group maintained the lumbar spine (Figure 4B) in extension, and the pelvis was more anteriorly tilted (Figure 4C) compared with the TFA-No LBP group (Figure 4C). The TFA-LBP group began reducing the anterior tilt after mid-stance, approximately 15% gait cycle earlier than the TFA-No LBP group (Figure 4C).

Persons with TTA

There were no notable differences between the TTA-LBP and TTA-No LBP groups in the coronal and sagittal plane (see Supplemental Digital Content,

In the transverse plane (Figure 5B), the lumbar segment of TTA-LBP group tended to rotate toward the intact side during late prosthetic stance as compared the TTA-No LBP group (Figure 5B). Although the mean curves seem to differ between the TTA-LBP and TTA-No LBP groups (Figure 5B), there was considerable overlap in the standard deviations that suggest a variable, but not different, motion pattern between these groups.

Figure 5:
Transverse plane kinematics of persons with TTA.


What do the results tell us about LBP in people with TFA?

In this study, differences in spinal and pelvic motion were observed between groups of people with TFA-LBP and TFA-No LBP in the coronal, transverse, and sagittal planes.

In the coronal plane, lateral trunk leaning toward prosthetic limb during single limb support is a common feature of TFA gait.9 It is thought to be a very effective way to position the center of mass (CoM) of the head/arms/trunk (HAT) close enough to the hip joint to reduce the demand on the gluteus medius musculature during prosthetic single support. There may be a number of reasons why reducing the demand on the gluteus medius is advantageous for people with TFA. Reducing the hip internal moment reduces the force driving the residual femur into the lateral wall of the socket which, in turn, may improve comfort on the lateral distal end of the cut femur and also reduce the socket interface pressures. It may serve as a way to reduce the coronal plane socket brim pressures necessary to stabilize the socket and residual limb interface. Perhaps some degree of lateral trunk lean is a requisite feature of TFA gait—irrespective of the presence/absence of LBP—given the demands on the socket/residuum interface that are reduced by lateral trunk leaning.

The results from our study suggest that, while the degree of lateral trunk lean is much the same in people with TFA-LBP and TFA-No LBP, there are different patterns of pelvic and lumbar motion used to achieve the lateral trunk lean observed. Perhaps those with TFA-LBP may not laterally flex the lumbar spine due to pain, or avoid using this strategy to protect the lumbar spine. Given the limited lateral lumbar flexion in the TFA-LBP group, it was necessary to elevate the pelvis on the intact side during swing—a hip-hiking pattern, effectively tipping the trunk toward the prosthetic limb to achieve the required degree of lateral trunk lean.

If these gait adaptations are used over a long period, it is likely that the lateral trunk muscles on the prosthetic side will become short,30 leading to changes in spinal alignment. The presence of such structural changes has been reported in persons with TFA and has been reasoned as a mechanical factor contributing to LBP.12 Given that the TFA-LBP group were middle-aged and with 25 more years of prosthetic use as compared with the TFA-No LBP group (Table 1), the TFA-LBP group may have developed long-term structural adaptations that may influence pain.

In the transverse plane, the motion of lumbar and thoracic spine toward the prosthetic side throughout the gait cycle is similar to that observed in people with scoliosis in the population of persons without amputation during gait,31 suggesting the potential influence of spinal alignment changes on LBP in people with TFA. Future researchers should consider measuring the trunk for alignment changes in evaluating subjects with TFA. This information would be valuable in understanding the influence of spinal alignment changes on LBP in people with TFA.

In the sagittal plane, the TFA-LBP group exhibited increased anterior pelvic tilt compared with the TFA-No LBP group. People in the TFA-LBP group may have greater hip flexion contracture compared with those without LBP, leading to compensations in the pelvic and lumbar spine that could affect their LBP. Shortness of the hip flexor musculature is a common complication after TFA30 because amputation leads to hip flexor/extensor muscle imbalance.32 Most people with TFA have no hip extension range and present with small hip flexion contractures.30 As the hip flexors become shorter, the anterior pelvic tilt increases because there just isn't the hip flexor muscle length to maintain a more neutral pelvic tilt. As anterior pelvic tilt increases, people may compensate by increasing lumbar lordosis to maintain the CoM of the HAT over the pelvis in standing and walking. Increased lumbar lordosis increases pressure on the posterior spinal structures (e.g., posterior disc, facet joints, nerve roots) and when maintained over an extended period may lead to chronic LBP.32 It is perhaps not surprising that the LBP group is considerably older with more years of prosthetic use than the No-LBP group. We might expect that people with TFA, like the population of persons without amputation, suffer more from muscle shortness as they age and have less spinal mobility with which to compensate. Unfortunately, we were not able to support this interpretation with detailed information on the available hip flexor muscle length (or degree of hip flexion contracture) and how well this was accommodated for design of the socket flexion angle. This information will be important for future researchers to gather to help inform this interpretation of the potential mechanism.

What do the results tell us about LBP in people with TTA?

In people with TTA, there were no differences in motion patterns in the coronal or sagittal plane between the LBP and No-LBP groups. In the transverse plane, people in the LBP group exhibited increased rotation of the lumbar spine toward the intact limb during prosthetic stance; consistent with other studies,26,28 but this increase was variable and small in magnitude. Previous studies26,28 have compared the spinal kinematics of persons with TTA and persons without amputation, and offer a poor basis for comparison.

How do results compare to other studies of spinal and pelvic motion in TFA gait?

The result from the present study differed from two previous studies.13,14

In the present study, transverse plane rotation of the lumbar spine was reduced in the TFA-LBP group compared with the TFA-No LBP group, whereas in the study by Morgenroth et al.,13 it was increased. Morgenroth et al.13 modeled the lumbar segment using markers placed in the lower thoracic spine (T8, T10, right and left of T10), and the present study modeled the lumbar segment using markers placed in the upper lumbar spine (L3, right and left of L3, T12). Although the movement characteristics of the lower thoracic and upper lumbar vertebrae are similar, the lower thoracic landmarks (T8 or T10) used in the Morgenroth et al. study13 are commonly recommended as optimal marker locations to investigate thoracic kinematics.33 Increased transverse rotation of the lumbar spine reported in that study13 is similar to the transverse rotations of thoracic segments of TFA-LBP and TFA-No LBP groups observed in the present study (Figure 2A).

Similarly, Fatone et al.14 found more variable lumbar movement patterns, which made it hard to identify motion patterns that could be attributable to the presence (or absence) of LBP. This could be partly attributed to differences in kinematic modelling strategies. For example, the present study used a marker triad approach (i.e., three markers placed over three distinct sites) using markers placed in the upper lumbar spine (L3, right and left of L3, T12) in contrast to a marker cluster (i.e., three noncollinear markers mounted on a stable base) placed over the L3 vertebra by Fatone et al.14 A recent kinematic study comparing marker triad versus marker clusters reported a considerable increase in mean absolute variability for lumbar transverse rotation using a marker cluster protocol (11°) as compared with marker triad protocol (4°).34 The results from lumbar kinematic studies using marker triad and marker cluster approaches may have to be interpreted with caution.

In terms of participant demographics, unlike the previous studies,13,14 the TFA-LBP and TFA-No LBP groups were not comparable (Table 1). The TFA-LBP group of the present study were older and with more years of prosthesis use as compared with the TFA-No LBP, which may have contributed to the differences in results from previous studies.

Limitations and Future Research

Descriptive studies, such as the present study, play an important role in emerging areas of research, given that it is prudent and ethical to undertake small studies that seek to identify gait adaptations that might be associated with LBP and identify measures that can be used to quantify these. These sorts of studies should lead to explanations of the rationale underlying causal mechanisms and the establishment of hypotheses that can be tested experimentally in future studies. Only in this way can experimental studies be hypothesis driven and well designed.

This investigation may help pave the way for experimental studies by identifying spinal and pelvic movement patterns that may differ between people with TFA-LBP or TFA-No LBP. Describing the potential LBP causal mechanisms in detail can guide the development of hypotheses. For example, our discussion suggests that shortness of the hip flexors leads to increased anterior pelvic tilt and compensatory increases in lumbar lordosis that, in turn, increase pressure on the posterior structures of the lumbar spine causing pain. Studies could be designed to treat the hip flexor tightness in people with TFA and evaluate the potential effect on LBP.

Although this study is limited by a small sample and a number of limitations of the method, the results offer novel insights that might help guide future research. By describing the limitations in detail, we hope to inform future research efforts.

We encourage other investigators to undertake a complete physical assessment that includes spinal and hip range of motion, muscle strength of spinal and hip musculature, as well as presence of postural asymmetries such as hip flexor and abductor muscle tightness/contractures and assessment of residual limb length.

Conventionally, the trunk is modeled as a single segment for assessing the spinal motion. However, evidences from in vitro studies suggest that regional variations exist concerning facet joint orientation and resulting regional spinal motion.35 Furthermore, a combination of movements from different spinal levels occurs during spinal motion (i.e., coupling).35 Therefore, to quantify spinal movement in more detail, the present study assessed the regional kinematics of spinal segments using a multi-segment kinematic model, including thoracic and lumbar segments in addition to a trunk segment. Future investigators should look to standardize the spinal marker placements to facilitate comparison between investigations.

Importantly, the study sample is not representative as only people with lower-limb amputation due to nonvascular etiology (i.e., trauma) were included. People with dysvascular amputation were excluded because of the confounding influence of systematic illnesses (e.g., cardiovascular problems, deficits in balance and muscle strength) that influence LBP. Future studies of LBP in people with TFA may have to assess the confounding influence of older age, more years with prosthesis use, presence of hip flexion contractures, residual limb length and prosthetic factors such as type of prosthetic knee or foot types, and prosthetic socket flexion angles. These factors could be used as covariants in inferential analyses comparing LBP in these groups. More representative groups may also include not only people with more representative age, sex distribution, and years of prosthesis use, but also people with a wide range of LBP severity, given that the participants from the present study, and previous studies on TFA,13,14 investigated people with TFA and mild LBP.

It is not possible to use the results of the present study to inform clinical practice given the small number of participants, particularly in the TFA groups. Furthermore, the TFA-LBP participants were pain free at the time of testing, suggesting the presence of longstanding spinal movement adaptations. It remains unclear whether the longstanding spinal movement adaptations are a sign of effective motor control strategy of the nervous system in preventing LBP36 or a “maladaptive” movement impairment contributing to ongoing LBP in this population.37 Therefore, future longitudinal research is needed to confirm these observations before clinical recommendations can be well informed.


This study provides preliminary evidence of differences in spinal and pelvic kinematics that may be associated with LBP in people with TFA. The TFA-LBP group elevated the pelvis on the intact side and minimized lumbar lateral flexion to achieve the same trunk lateral flexion angle (Trendelenburg gait) as those with TFA-No LBP. Those with TFA-LBP kept the lumbar spine rotated toward the prosthetic limb throughout the gait cycle. There were no differences in the spinal and pelvic kinematics of people with TTA-LBP and TTA-No LBP.


We wish to thank NZALS for their assistance with study recruitment.


1. Gailey R, Allen K, Castles J, et al. Review of secondary physical conditions associated with lower-limb amputation and long-term prosthesis use. J Rehabil Res Dev 2008;45:15–29.
2. Devan H, Tumilty S, Smith C. Physical activity and lower-back pain in persons with traumatic transfemoral amputation: a national cross-sectional survey. J Rehabil Res Dev 2012;49:1457–1466.
3. Hoy D, Bain C, Williams G, et al. A systematic review of the global prevalence of low back pain. Arthritis Rheum 2012;64:2028–2037.
4. Smith DG, Ehde DM, Legro MW, et al. Phantom limb, residual limb, and back pain after lower extremity amputations. Clin Orthop Relat Res 1999:29–38.
5. Ehde DM, Smith DG, Czerniecki JM, et al. Back pain as a secondary disability in persons with lower limb amputations. Arch Phys Med Rehabil 2001;82:731–734.
6. Ephraim PL, Wegener ST, MacKenzie EJ, et al. Phantom pain, residual limb pain, and back pain in amputees: results of a national survey. Arch Phys Med Rehabil 2005;86:1910–1919.
7. Taghipour H, Moharamzad Y, Mafi AR, et al. Quality of life among veterans with war-related unilateral lower extremity amputation: a long-term survey in a prosthesis center in Iran. J Orthop Trauma 2009;23:525–530.
8. da Costa BR, Vieira ER. Risk factors for work‐related musculoskeletal disorders: a systematic review of recent longitudinal studies. Am J Ind Med 2010;53:285–323.
9. Devan H, Carman A, Hendrick P, et al. Spinal, pelvic, and hip movement asymmetries in people with lower-limb amputation: systematic review. J Rehabil Res Dev 2015;52:1–19.
10. Devan H, Carman AB, Hendrick PA, et al. Perceptions of low back pain in people with lower limb amputation: a focus group study. Disabil Rehabil 2015;37:873–883.
11. Hoogendoorn WE, van Poppel MN, Bongers PM, et al. Physical load during work and leisure time as risk factors for back pain. Scand J Work Environ Health 1999;25:387–403.
12. Devan H, Hendrick P, Ribeiro DC, et al. Asymmetrical movements of the lumbopelvic region: is this a potential mechanism for low back pain in people with lower limb amputation? Med Hypotheses 2014;82:77–85.
13. Morgenroth DC, Orendurff MS, Shakir A, et al. The relationship between lumbar spine kinematics during gait and low-back pain in transfemoral amputees. Am J Phys Med Rehabi 2010;89:635–643.
14. Fatone S, Stine R, Gottipati P, et al. Pelvic and spinal motion during walking in persons with transfemoral amputation with and without low back pain. Am J Phys Med Rehabil 2016;95:438–447.
15. Dionne CE, Dunn KM, Croft PR, et al. A consensus approach toward the standardization of back pain definitions for use in prevalence studies. Spine (Phila Pa 1976) 2008;33:95–103.
16. Roland M, Morris R. A study of the natural history of back pain: part I: development of a reliable and sensitive measure of disability in low-back pain. Spine 1983;8:141–144.
17. Gauthier-Gagnon C, Grisé MC. Tools to measure outcome of people with a lower limb amputation: update on the PPA and LCI. J Prosthet Orthot 2006;18:61–67.
18. Hanspal R, Fisher K, Nieveen R. Prosthetic socket fit comfort score. Disabil Rehabil 2003;25:1278–1280.
19. Bjelland I, Dahl AA, Haug TT, et al. The validity of the Hospital Anxiety and Depression Scale. An updated literature review. J Psychosom Res 2002;52:69–77.
20. Intolo P, Carman AB, Milosavljevic S, et al. The Spineangel®: examining the validity and reliability of a novel clinical device for monitoring trunk motion. Man Ther 2010;15:160–166.
21. Wu G, Siegler S, Allard P, et al. ISB recommendation on definitions of joint coordinate system of various joints for the reporting of human joint motion–part I: ankle, hip, and spine. International Society of Biomechanics. J Biomech 2002;35:543–548.
22. Visual 3D. Visual 3D Pelvis. 2014. Available at:
23. Visual 3D. Visual 3D Thorax/Abdomen. 2014. Available at:
24. Leardini A, Biagi F, Belvedere C, et al. Quantitative comparison of current models for trunk motion in human movement analysis. Clin Biomech (Bristol, Avon) 2009;24:542–550.
25. Goujon-Pillet H, Sapin E, Fodé P, et al. Three-dimensional motions of trunk and pelvis during transfemoral amputee gait. Arch Phys Med Rehabil 2008;89:87–94.
26. Hendershot BD, Wolf EJ. Three-dimensional joint reaction forces and moments at the low back during over-ground walking in persons with unilateral lower-extremity amputation. Clin Biomech (Bristol, Avon) 2014;29:235–242.
27. Esposito ER, Wilken JM. The relationship between pelvis–trunk coordination and low back pain in individuals with transfemoral amputations. Gait Posture 2014;40:640–646.
28. Yoder AJ, Petrella AJ, Silverman AK. Trunk–pelvis motion, joint loads, and muscle forces during walking with a transtibial amputation. Gait Posture 2015;41:757–762.
29. Roland M, Fairbank J. The Roland–Morris disability questionnaire and the Oswestry disability questionnaire. Spine 2000;25:3115–3124.
30. Gaunaurd I, Gailey R, Hafner BJ, et al. Postural asymmetries in transfemoral amputees. Prosthet Orthot Int 2011;35:171–180.
31. Yang JH, Suh SW, Sung PS, et al. Asymmetrical gait in adolescents with idiopathic scoliosis. Eur Spine J 2013;22:2407–2413.
32. Friel K, Domholdt E, Smith DG. Physical and functional measures related to low back pain in individuals with lower-limb amputation: An exploratory pilot study. J Rehabil Res Dev 2005;42:155–166.
33. Armand S, Sangeux M, Baker R. Optimal markers’ placement on the thorax for clinical gait analysis. Gait Posture 2014;39:147–153.
34. Kiernan D, Malone A, O’Brien T, et al. A quantitative comparison of two kinematic protocols for lumbar segment motion during gait. Gait Posture 2015;41:699–705.
35. Panjabi M, Yamamoto I, Oxland T, et al. How does posture affect coupling in the lumbar spine? Spine 1989;14:1002–1011.
36. Hodges PW, Moseley GL. Pain and motor control of the lumbopelvic region: effect and possible mechanisms. J Electromyogr Kinesiol 2003;13:361–370.
37. O’Sullivan P. Diagnosis and classification of chronic low back pain disorders: maladaptive movement and motor control impairments as underlying mechanism. Man Ther 2005;10:242–255.

amputation; kinematics; low back pain; musculoskeletal; spine; walking

Supplemental Digital Content

Copyright © 2017 American Academy of Orthotists and Prosthetists