SPINE Volume 26, Number 14, pp E330–E337
©2001, Lippincott Williams & Wilkins, Inc.
Prospective Dynamic Functional Evaluation of Gait and
Spinal Balance Following Spinal Fusion in Adolescent
Idiopathic Scoliosis
Lawrence G. Lenke, MD,* Jack R. Engsberg, PhD,† Sandy A. Ross, MHS,†
Angela Reitenbach, MHS,† Kathy Blanke, RN,* and Keith H. Bridwell, MD*
Study Design. Prospective evaluation of gait and spinal–pelvic balance parameters in patients with adolescent
idiopathic scoliosis undergoing a spinal fusion.
Objective. To evaluate changes in gait and three-dimensional alignment and balance of the spine relative to
the pelvis as a consequence of spinal fusion.
Summary of Background Data. Preoperative and postoperative spinal radiographs have been the major forms
of outcome analysis of adolescent idiopathic scoliosis
fusions. The use of optoelectronic analysis for posture
and gait has gained acceptance recently. However, there
is a paucity of data quantifying, comparing, and correlating structural and functional changes in patients undergoing scoliosis fusion surgery including upright posture
and gait.
Methods. Thirty patients with adolescent idiopathic
scoliosis undergoing an instrumented spinal fusion
were prospectively evaluated. Coronal and sagittal vertical alignment was evaluated on radiographs (CVA-R,
SVA-R), during upright posture (CVA-P and SVA-P), and
during gait (CVA-G, SVA-G). Transverse plane alignment was evaluated by the acromion–pelvis angle during gait.
Results. Gait speed was significantly decreased (P ⬍
0.05) between preoperative (129 ⫾ 16 cm/sec) and 2-year
postoperative (119 ⫾ 16 cm/sec) testing sessions. Decreasing gait speed was the result of significantly reduced
cadence and decreased stride length. There were no significant differences for lower extremity kinematics over
the entire gait cycle. Spinal–pelvic balance parameters
showed significant improvement in mean CVA-R, CVA-G
(P ⬍ 0.05), then unchanged CVA-P at 2 years postoperation. CVA-P was relatively unchanged while the mean
CVA-G also showed significant improvement from preoperation (2.2 ⫾ 2.4 cm) to 2 years postoperation (1.3 ⫾ 1.3
cm)(P ⬍ 0.05). The mean SVA-R, SVA-P, and SVA-G were
unchanged at 2 years postoperation (P ⬎ 0.05). The acromion–pelvis angle during gait at maximum shoulder rotation was statistically improved at 1 year (P ⫽ 0.002) and
2 years (P ⫽ 0.001) after surgery. Importantly, CVA-P correlated with CVA-G, and SVA-P correlated with SVA-G to
the P ⬍ 0.05 level.
Conclusions. Patients with adolescent idiopathic scoliosis undergoing spinal fusion show slightly decreased
gait speed at 2 years postoperation without any change in
From the *Department of Orthopaedic Surgery, Washington University School of Medicine, and †Human Performance Laboratory, Barnes-Jewish Hospital, St. Louis, Missouri.
Presented at the Scoliosis Research Society 33rd Annual Meeting, New
York, New York, September 1998.
Acknowledgment date: October 10, 2000.
First revision date: January 3, 2001.
Acceptance date: March 29, 2001.
Device status category: 11.
Conflict of interest category: 15.
E330
lower extremity kinematics. Spinal–pelvic balance parameters are improved in the coronal plane and unchanged in
the sagittal plane radiographically and during standing
posture and gait. Transverse plane parameters also are
improved at maximum shoulder rotation during gait. [Key
words: adolescent idiopathic scoliosis, coronal balance,
sagittal balance, optoelectronic analysis, spinal–pelvic
balance, gait] Spine 2001;26:E330 –E337
Spinal fusions are still the primary means of correcting a
scoliosis deformity and thereby halting progression. The
most common patient to have this kind of surgery is an
adolescent who has idiopathic scoliosis12,17 and whose
etiology is still unknown.5 Frequently, this fusion extends from the thoracic region into varying portions of
the lumbar spine and can be performed via the posterior,
anterior, or combined route. Although partial correction
of the curvature is obtained, it is at the expense of removing normal intervertebral motions that exist in the scoliotic spine.19
The preoperative analysis and the postoperative documentation of surgical results are currently performed
via long cassette radiographs of the spinal column in the
coronal and sagittal projection. There are broad radiographic goals for scoliosis correction in the coronal plane
and they include the following: 1) a balanced spine with
a plumb line from C7 bisecting the sacrum; 2) a balanced
thoracic and lumbar spine with the top and bottom of the
instrumentation and fusion falling within the stable zone
of Harrington; and 3) for fusions into the mid and lower
lumbar spine (L2 and below) that the lowest instrumented vertebra (LIV) be horizontal to the pelvis, bisected by the center sacral line and neutrally rotated.13,18,21 In the sagittal plane the radiographic goals are
as follows: 1) a balanced sagittal plane with the plumb
line from C7 falling at/or behind the lumbosacral disc; 2)
normal thoracic, thoracolumbar, and lumbar sagittal
plane alignment within the instrumented and uninstrumented regions; and 3) to have junctions between the
regional segments of thoracic kyphosis and lumbar lordosis neutral while avoiding a junctional kyphosis.21
Since the advent of spinal instrumentation as an adjunct to operative correction and fusion, radiographs
have been the major forms of outcome analysis of scoliosis fusions. With the current use of segmental spinal
instrumentation systems, three-dimensional analysis of
the deformity preoperation and postoperation has become important. This has improved radiographic results,
Adolescent Idiopathic Scoliosis • Lenke et al E331
Table 1. Demographic and Surgical Data for 30 Patients
With AIS
Patient
Number
Patient
Age
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
16
18 ⫹ 8
12 ⫹ 8
18 ⫹ 1
12 ⫹ 7
12 ⫹ 8
15 ⫹ 7
13 ⫹ 3
13 ⫹ 7
17 ⫹ 7
14 ⫹ 11
13 ⫹ 2
15 ⫹ 2
17 ⫹ 11
13 ⫹ 11
14 ⫹ 6
11 ⫹ 6
15 ⫹ 2
13 ⫹ 9
14 ⫹ 9
14 ⫹ 11
14 ⫹ 2
17 ⫹ 1
14 ⫹ 6
13 ⫹ 2
18 ⫹ 5
13 ⫹ 7
18 ⫹ 8
13 ⫹ 2
16 ⫹ 9
Sex
Type of Fusion
ASF/PSF/Combo
Total Number of
Fusion Levels
LIV
F
F
F
F
F
M
F
F
M
F
F
F
F
M
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
PSF
PSF
PSF
PSF
PSF
PSF
PSF
PSF
PSF
PSF
PSF
PSF
PSF
PSF
PSF
PSF
ASF
ASF
ASF
ASF
ASF
ASF
ASF
ASF
ASF
ASF
ASF
ASF
ASF/PSF
ASF/PSF
T4–L2 (10)
T2–L1 (11)
T2–L1 (11)
T4–L2 (10)
T4–T12 (8)
T4–L4 (12)
T3–L3 (12)
T4–L1 (9)
T3–L2 (11)
T2–L2 (12)
T5–L4 (13)
T4–T12 (8)
T4–L3 (11)
T4–L1 (9)
T3–L3 (12)
T3–T12 (9)
T5–T11 (6)
T4–T11 (7)
T5–T12 (7)
T11–L3 (4)
T5–T11 (6)
T5–L1 (8)
T5–T12 (7)
T10–L3 (5)
T6–L1 (7)
T5–T11 (6)
T5–T13 (8)
T5–T11 (6)
T2–L3 (13)
T12–L2 (2)
L2
L1
L1
L2
T12
L4
L3
L1
L2
L2
L4
T12
L3
L1
L3
T12
T11
T11
T12
L3
T11
L1
T12
L3
L1
T11
T13
T11
L3
L2
LIV ⫽ lowest instrumented vertebrae.
especially with increased emphasis placed on the sagittal
plane.13,21 Results for correction of axial plane deformity have been less successful.1 Nevertheless, although
changes in static alignment of the spinal column as assessed by standard radiographs do quantify changes in
structure, they do not quantify changes in function from
a dynamic perspective. A paucity of data exists quantifying, comparing, and correlating structural and functional changes in patients undergoing scoliosis fusion
surgery. The purpose of this study is twofold: 1) to evaluate changes in gait as a consequence of spinal fusion in
adolescents with idiopathic scoliosis and 2) to evaluate
changes in the three-dimensional alignment and balance
of the spine relative to the pelvis during standing posture
and gait. To this end we hope to gain valuable insight as
to the functional effects of a scoliosis fusion from a more
dynamic perspective.
Materials and Methods
This study details 30 patients (28 female and two male) with
adolescent idiopathic scoliosis (AIS) who underwent an instrumented spinal fusion to treat their scoliosis. The mean age of
the patients was 14 years range 12–18 years), and all were
treated with one of three types of surgical procedures involving
a posterior spinal fusion (PSF, n ⫽ 16), an anterior spinal fusion (ASF, n ⫽ 12), or a combined ASF/PSF (n ⫽ 2). All patients
treated with a PSF had posterior segmental spinal instrumentation placed for correction of deformity and performance of a
spinal fusion using autogenous bone harvested from the iliac
crest or the convex ribs to improve any associated rib prominence. Patients with an ASF had a single anterior screw rod
instrumentation augmented with rib autograft placed in the
disc spaces for the anterior fusion. Patients with a combined
ASF/PSF had anterior release and autogenous rib fusion performed followed by a PSF procedure with instrumentation as
described above. The demographic patient data along with the
type of surgery performed, total amount of fusion levels, and
the lowest instrumented vertebrae are provided for each patient
in Table 1. The mean number of levels fused was 10.5 (range
8 –13). The LIV was T11 (n ⫽ 5), T12 (n ⫽ 5), T13 (n ⫽ 1), L1
(n ⫽ 6), L2 (n ⫽ 5), L3 (n ⫽ 6), and L4 (n ⫽ 2).
All patients had preoperative, 1-year, and 2-year postoperative long cassette spinal radiographs performed in the coronal
and sagittal planes. All coronal curves were measured by the
Cobb method, and radiographic coronal balance (i.e., coronal
vertical alignment [CVA-R])18 was measured as a C7 plumb
line distance in centimeters from the midsacral line. Cobb measurements also were made in the sagittal plane measuring thoracic kyphosis (T5–T12), lumbar lordosis (T12–S1), and overall radiographic sagittal balance (i.e., sagittal vertical alignment
[SVA-R])18 as a C7 plumb line measured in centimeters as the
distance from the posterior–superior edge of S1. A positive
value indicated that the C7 plumb line fell in front of the posterior superior aspect of S1, whereas a negative value indicated
that this line fell in back of this point.24
To prepare the subject for the videographic gait analysis, 36
retroreflective surface markers were placed at selected locations
on segments of the lower extremities, trunk, and head of each
patient. A subset of 24 markers were used in the analysis of the
present investigation (Table 2). For the collection of the gait
data, each subject walked barefoot along a 9-m walkway and
had video data from a 6 camera HiRes Motion Analysis Corporation System (Santa Rosa, CA) collected during the middle
2 m (Figure 1). Five to seven trials of data were collected from
each subject. In addition, one trial of posture data was collected
while the patient stood in a relaxed manner.
For the posture and gait trials of each subject, the surface
marker location data were converted to three-dimensional coordinates as a function of time, then tracked and edited to
Table 2. Surface Markers Used to Represent Body Segments or Specific Vertebrae
Segment or Vertebrae
Shoulders
Vertebrae
Pelvis
Thighs
Legs
Feet
1st Surface Marker
2nd Surface Marker
Right acromion process
C7
T10
Right ASIS
Greater trochanter
Fibular head
Superior navicular
Left acromion process
S2
L4
Left ASIS
Mid anterior portion of femur
Mid anterior portion of tibia
Posterior calcaneus
3rd Surface Marker
S2
Lateral epicondyle of femur
Lateral malleolus
5th metatarsal head
332 Spine • Volume 26 • Number 14 • 2001
Figure 3. Mean acromion pelvis angle (APA-G) during the gait
cycle showing minimum degrees of difference occurring just after
the 20% of the gait cycle, with the maximum greatest range of
motion seen just past the halfway point of the gait cycle.
Figure 1. Patient with preoperative scoliosis undergoing the optoelectronic video gait analysis.
produce quantified measurements of standing posture and selected gait parameters. The standing posture measurements
were coronal vertical alignment (CVA-P) and sagittal vertical
alignment (SVA-P). These quantify the horizontal distance
from S2 to a vertical line dropped from C7 in the coronal and
sagittal planes, respectively. They were designed to be similar
to the previously described CVA-R and SVA-R radiographic
parameters.
The gait data were analyzed to quantify trunk alignment,
limb motion, and basic gait parameters. Seven distinct analyses
were performed. The first four were the coronal vertical alignment (CVA-G) and the sagittal vertical alignment (SVA-G), for
both right and left steps. Specifically, the CVA-G and SVA-G
were defined as above, but the values were from initial contact
during gait of the right and left feet. The initial contact time was
selected to represent the entire gait cycle because little change
was observed in the values over the cycle (Figure 2).
The final three gait parameters in the first set were related to
the transverse plane motion of the two acromion process markers relative to the two pelvis ASIS markers. Specifically, the
acromion–pelvis angle during gait (APA-G) was the transverse
plane angle created by the line formed from the right and left
acromion process markers, relative to the line created by the
two ASIS markers. A positive APA-G value was calculated if
the right acromion process marker was oriented anterior to the
right ASIS marker (i.e., protracted right shoulder), and a negative value was calculated if the right acromion process was
oriented posterior to the right ASIS marker (i.e., retracted right
shoulder). The first of the three parameters was the maximum
APA-G (APA-G Max)(Figure 3). The second was the minimum
APA-G (APA-G Min). The third parameter was the range of
motion over the entire gait cycle in the transverse plane and was
defined to be the difference between the maximum and minimum APA values (APA-G ROM).
We also analyzed basic gait parameters including gait speed,
stride length, cadence, and step width.7 These were analyzed
from a subset of 10 of the 30 original patients. This analysis
first included a visual inspection of the ankle, knee, and hip
joints as a function of the gait cycle in the sagittal, coronal, and
Figure 2. Mean coronal (CVA-G) and sagittal (SVA-G) spinal balance during the entire gait cycle.
Adolescent Idiopathic Scoliosis • Lenke et al E333
Figure 4. Lower extremity kinematic data including right knee flexion– extension, right angle inversion– eversion, right ankle dorsiflexion–
plantarflexion, and right foot rotation during the entire gait cycle showing minimal changes from preoperation to 1–2 years postoperation.
There are slight differences in the able bodied (AB) patient data.
transverse planes for the three test sessions (preoperative and
12 and 24 months postoperative). Because little variation was
observed across test sessions, four representative joint angles
over the gait cycle in a single plane were chosen to represent the
lower extremity gait changes (Figure 4). These were as follows:
1) right sagittal plane knee angle, 2) right sagittal plane ankle
angle, 3) right coronal plane ankle angle (eversion–inversion),
and 4) right transverse plane foot orientation (foot progression
angle). From each curve the numerical value at initial contact
was selected for use in the statistical analysis.
The statistical analysis included a multiple analysis of variance to determine any differences among the preoperative and
two postoperative test sessions. In addition, a Pearson Product
Moment Correlation Coefficient (r) was used to determine
whether respective relationships existed among the CVA and
SVA variables.
Results
Radiographic
The radiographic results are listed in Table 3. The mean
preoperative thoracic Cobb measurement was 57° (range
48 –77°), decreasing to a mean 27° (range 13– 40°) postoperation, for an average 53% instrumented correction
(range 19 –77%). The radiographic coronal plumb line
(CVA-R) improved from a mean 0.76 cm off balance
preoperative (range 0 –5.9 cm) to 0.01 cm (range 0 –3.3
cm) at 1 year postoperation, and 0.21 cm (range 0 –3.0
cm) at 2 years postoperation (P ⬍ 0.05). There were no
significant changes noted in the radiographic and postural sagittal data at 2 years postoperation (P ⬎ 0.05).
Table 3. Radiographic Data for All 30 Patients
Sagittal Alignment
Cobb
Patient
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
CVA-R
T5–T12
L1–Sac
SVA-R
Pre
2 yr
Pre
1 yr
2 yr
Pre
2 yr
Pre
2 yr
Pre
1 yr
2 yr
53
55
77
57
57
48T/58L
45T/61L
56
62
60
55T/62L
61
53
52
53
48
59
63
69
34
57
82
40
22T/49TL
66
50
66
52
82
42TL
57°
19
22
40
15
35
33T/30L
18T/23L
25
28
47
36T/27L
42
13
27
38
39
33
35
35
16
18
37
23
26T/28TL
23
38
35
25
28
19
27° MEAN
0.8R
1.8L
2.5L
2.3R
4L
2.6R
1.9R
0.4L
1.2R
1.1R
0.9R
2.6L
1.6R
2.6L
2.3R
0
1.5R
1.6R
0.4R
2.9R
0.7L
0.9R
5.9R
4.7R
0.9R
3.9R
1.6L
2R
2R
2.4R
0.75926
0.4L
0.31L
1.5L
1L
1.6L
2.1R
0
1.1L
0.1R
0.1R
0.3R
3.3L
3R
0.1R
0.1R
2R
3.2L
1.5L
2.3R
2.2R
1.0R
0.1R
2.9R
2.3R
0.2L
2.4L
2.5L
2.2L
0.6R
1.4R
0.0107
0.9L
0.3R
0
0.3R
3L
0.8R
1.2R
1.5L
0.2R
2.2R
0
2.2L
1.8R
0.8L
0.2R
0
1.3L
1.2L
0
0.4R
0.3L
1.9R
1R
0.1R
0.2R
1.2L
1.4L
3L
0.2R
0.9
0.2077
36
30
43
25
39
18
23
29
24
37
36
34
41
30
46
16
28
13
19
14
4
20
30
22
9
45
58
⫺5
35
48
28.071
20
28
29
7
24
22
22
23
16
43
40
18
30
15
28
20
31
20
17
29
23
31
27
22
27
46
55
10
27
46
26.852
⫺61
⫺65
⫺60
⫺62
⫺64
⫺42
⫺55
⫺79
⫺57
⫺58
⫺81
⫺48
⫺84
⫺82
⫺71
⫺54
⫺76
⫺66
⫺63
⫺55
⫺66
⫺72
⫺74
⫺59
⫺58
⫺59
⫺66
⫺59
⫺73
⫺65
⫺64.46667
⫺52
⫺62
⫺53
⫺56
⫺61
⫺58
⫺70
⫺71
⫺47
⫺68
⫺74
⫺52
⫺68
⫺72
⫺72
⫺58
⫺60
⫺72
⫺57
⫺70
⫺63
⫺69
⫺68
⫺59
⫺56
⫺70
⫺68
⫺50
⫺60
⫺69
⫺62.407
⫺8.1
⫺0.9
⫺2.2
⫺7.2
⫹1.9
⫺4.4
⫺1.1
⫺3
⫹4.5
⫹.4
⫺9.3
⫹4.9
⫺2.3
⫹0.8
⫺1.3
⫹0.3
⫺8.4
⫹3.3
⫺4.8
⫺0.7
0.6
⫺5.2
⫺0.3
⫺2.1
⫹0.9
⫹4.1
⫺5.1
⫺2.7
⫺3.3
⫺2.1
⫺3.519
⫺0.2
⫺0.9
⫹3.3
⫺6
⫺0.8
⫺3.8
⫺1.2
⫺2.5
⫺3.6
⫺4.5
⫺4.1
⫹0.9
⫹0.4
⫹3.6
⫺7.7
⫺2.8
⫹1.7
⫺3.4
⫹3.4
⫺4.2
⫺4.1
⫺7.6
⫺1.1
⫺0.9
⫹1.3
⫺2.6
⫺1
⫺1.6
⫺6
⫺2.7
⫺3.145
⫹3.2
⫹0.5
⫹1.2
⫺2.5
⫺1
⫹1.7
⫺0.7
⫺5.6
⫹1.7
⫺4.8
0
⫹1.7
⫹1.1
⫹1.6
⫺5.6
⫺4.1
⫺6.4
⫹1.1
0
⫺4.2
⫺0.4
⫺1.6
⫹0.1
⫹1.5
⫹1.7
⫹1.5
1.6
⫺0.4
⫺6.3
⫺3.2
⫺3.0
E334 Spine • Volume 26 • Number 14 • 2001
Gait Analysis
Gait speed was significantly decreased (Table 4) between
the preoperative test session (129 ⫾ 16 cm/sec) and the
two postoperative sessions (12 months, 120 ⫾ 16 cm/
sec; 24 months, 119 ⫾ 16 cm/sec). The decrease in gait
speed was a result of a significantly reduced cadence for
both comparisons (Table 4) and a significantly decreased
stride length for the preoperative to 24-month postoperative test sessions. There were no differences between the
two postoperative sessions for speed, cadence, and stride
length. There was a significant decrease in stride width
between the preoperative condition and the 12-month
postoperative condition, but not the 2-year postoperative condition. There was no stride width difference between the two postoperative conditions, and no difference in hip joint/pelvic motion.
Using the value for initial contact (0%) of the gait
cycle for each joint measurement (Figure 4) for the subgroup of 10 subjects indicated there were no significant
differences between test sessions preoperation, 1 year,
and 2 years postoperation (Table 5). Thus, the results
showed no differences in the lower extremity kinematic
data for these 10 patients, and no further analysis was
performed.
Spinal–Pelvic Alignment and Balance
Coronal plane spinal–pelvic balance during upright posture (CVA-P) averaged ⫺0.5 ⫾ 2.3 cm preoperation,
0.4 ⫾ 1.1 at 1 year postoperation, and 0.0 ⫾ 1.3 cm at 2
years postoperation. One-year but not 2-year postoperative values were statistically different from the preoperative values (P ⬍ 0.05). CVA-G at right initial foot contact had a mean preoperative value of 2.3 ⫾ 2.4 cm,
decreasing to 1.4 ⫾ 1.4 cm at 1 year postoperation, and
decreasing slightly again to 1.3 ⫾ 1.3 cm at 2 years postoperation. The differences between preoperative and 1
year (P ⫽ 0.028) and between the preoperative and
2-year postoperative values (P ⫽ 0.009) were both statistically significant. Interestingly, right CVA-G was significantly improved while left CVA-G was not. This may
reflect the fact that all patients included in this series have
a curvature that was convex to the right (right thoracic).
There is no statistical difference between the SVA-P
preoperation at 1 year and 2 years postoperation (P ⬎
0.05). Similarly, for both SVA-G at right and left initial
foot contact, there were no significant differences between the preoperative, 1-year, and 2-year postoperative
values (P ⬎ 0.05) (Table 6).
In the transverse plane APA-G Min showed no significant differences between preoperative and 2-year postoperative values (P ⬎ 0.05). The APA was also calculated at maximum shoulder rotation (APA-G Max) with
a preoperative mean of ⫺12.5 ⫾ 5.1°, decreasing to
⫺8.9 ⫾ 5.1 at 1 year postoperation, and decreasing
slightly to ⫺7.1 ⫾ 3.6 at 2 years postoperation. The
preoperative APA at maximum shoulder rotation was
statistically significant at the 1-year and 2-year postoperative values (P ⫽ 0.002 and 0.001, respectively).
Lastly, the APA-G ROM preoperative showed a mean
value of 14.2 ⫾ 4.1°, decreasing to 10.2 ⫾ 2.9° at 1 year
postoperation and decreasing slightly again to 9.3 ⫾ 3.5°
at 2 years postoperation. The values between the preoperative and 1-year preoperative and 2-year postoperative
measurements were both highly statistically significant
(P ⬍ 0.0001).
In summary, the SVA-G did not show any differences
between preoperative and postoperative values during
both right and left initial foot contact. However, the
CVA-G at right initial foot contact showed statistically
significant differences with improved coronal balance at
both 1 year and 2-year postoperation compared with
preoperation. Interestingly, at left initial foot contact,
these changes were not statistically significant. Finally,
using the transverse position of the shoulders relative to
the pelvis as determined by the APA-G, highly significant
changes were noted between preoperative and postoperative values. Specifically, there was an improved angular
symmetry found in the maximum transverse angles
(APA-G Max) during the gait cycle. We believe this indicated an improvement occurred because the shoulders
rotated toward a more equal angular alignment with the
pelvis postoperative. However, the decline in transverse
plane range of motion (APA-G ROM) could be a limitation of the surgery.
There was no correlation between the radiographic
data and the postural data (CVA-P and SVA-P) as well as
the gait data (CVA-G and SVA-G) for the coronal or
sagittal plane alignment assessment. However, a positive
correlation was seen between the sagittal standing postural data (SVA-P) to the gait data (SVA-G) (P ⬍ 0.05).
In addition, there was an extremely strong correlation
between the coronal standing postural data (CVA-P) and
the gait data (CVA-G) (P ⬍ 0.01). This strong correlation
was also noted between the right (CVA-G and SVA-G) and
left (CVA-G and SVA-G) initial foot contact coronal and
sagittal plane gait data (P ⬍ 0.01).
Discussion
Table 4. Linear Gait Data (n ⴝ 30)
Speed (cm/s)
Cadence (steps/min)
Stride length (cm)
Stride width (cm)
Preop
12 Month Postop
24 Month Postop
129 ⫾ 16
120 ⫾ 8
128 ⫾ 11
8.1 ⫾ 3.0
120 ⫾ 16*
115 ⫾ 8*
125 ⫾ 11
7.2 ⫾ 2.7*
119 ⫾ 16*
114 ⫾ 9*
124 ⫾ 12*
7.4 ⫾ 2.2
* Significantly different from preop (P ⬍ 0.05).
Standing long cassette coronal and sagittal radiographs
are the standard means of preoperative analysis and
postoperative assessment of the surgical results of scoliosis fusions.12 Although changes in the static alignment
of the spinal column, as is assessed by standard radiographs, do quantify changes in structure, they do not
address concomitant changes in patient function that
may occur. Currently, a paucity of data exists quantify-
Adolescent Idiopathic Scoliosis • Lenke et al E335
Table 5. Kinematic Gait Data (n ⴝ 10)
Sagittal knee angle at initial contact
Sagittal ankle angle at initial contact
Coronal ankle angle at initial contact
Transverse foot angle at initial contact
Preop Degrees
12 Month Postop
Degrees
24 Month Postop
Degrees
3.3 ⫾ 1.0
⫺3.0 ⫾ 3.2
3.1 ⫾ 2.4
⫺11.7 ⫾ 4.6
3.8 ⫾ 1.1
⫺2.0 ⫾ 3.2
3.4 ⫾ 2.9
⫺10.8 ⫾ 4.5
2.8 ⫾ 0.9
⫺3.0 ⫾ 2.3
3.1 ⫾ 2.8
⫺11.5 ⫾ 2.7
ing structural and functional changes in patients undergoing scoliosis fusion surgery.11 This study attempts to
address how scoliosis fusion affects gait and spinal–
pelvic balance alignment parameters after surgery using
optoelectronic techniques. There have been a few studies
analyzing various gait parameters in scoliotic patients.
Thurston and Harris were the first to analyze normal
kinematics of the lumbar spine and pelvis using gait analysis in 48 male volunteers ranging in age from 16 to 74
years.23 They presented results for both range of motion
in each plane as well as a wave pattern for movement in
each plane over time. Analysis of the wave pattern
showed that movements of the pelvis and lumbar spine
relate directly to identifiable events in the gait cycle.
Giakas et al analyzed gait patterns between healthy
and patients with scoliosis via analysis of ground reaction forces.9 They concluded that patients with scoliosis
exhibited balance problems during the stance phase of
gait and have significant asymmetry in the frequency
characteristics of gait. No kinematics were reported.
Hopf et al studied idiopathic scoliosis patients before
and after surgery to evaluate muscle activation patterns.14 They analyzed 23 patients before and after posterior instrumentation and fusion with truncal and lower
extremity EMGs. The results supported the hypothesis
that activity asymmetries observed in the paravertebral
musculature in patients with idiopathic scoliosis are the
result of the scoliotic body deformities with consequent
asymmetries in the biomechanical force patterns of body
Table 6. Spinal-Pelvic Balance Parameters
CVA-R (cm)
SVA-R (cm)
CVA-P (cm)
SVA-P (cm)
Right CVA-G (cm)
Left CVA-G (cm)
Right SVA-G (cm)
Left SVA-G (cm)
APA-G Max (degrees)
APA-G Min (degrees)
APA-G ROM (degrees)
Preop
1 Yr Postop
2 Yr Postop
0.8 ⫾ 2.4
⫺3.52 ⫾ 3.8
⫺0.5 ⫾ 2.3
4.6 ⫾ 2.6
2.3 ⫾ 2.4
⫺0.0 ⫾ 1.9
6.1 ⫾ 2.6
6.2 ⫾ 2.6
12.5 ⫾ 5.1
⫺1.7 ⫾ 4.5
14.2 ⫾ 4.1
0.0 ⫾ 1.8
⫺3.15 ⫾ 3.0
0.4 ⫾ 1.1*
4.5 ⫾ 2.4
1.4 ⫾ 1.4*
⫺0.3 ⫾ 1.3
5.8 ⫾ 2.6
5.8 ⫾ 2.6
8.9 ⫾ 5.1*
⫺1.3 ⫾ 4.6
10.2 ⫾ 2.9*
⫺0.2 ⫾ 1.3*
⫺3.0 ⫾ 2.9
0.0 ⫾ 1.3
5.0 ⫾ 2.6
1.3 ⫾ 1.3*
⫺0.4 ⫾ 1.3
6.4 ⫾ 3.1†
6.4 ⫾ 2.8†
7.1 ⫾ 3.6* †
⫺2.2 ⫾ 4.3
9.3 ⫾ 3.5*
* Significantly different from preop (P ⬍ 0.05).
† Significantly different from 12 month postop (P ⬍ 0.05).
CVA-R ⫽ coronal vertical alignment–radiographic; SVA-R ⫽ sagittal vertical
alignment–radiographic; CVA-P ⫽ coronal vertical alignment–postural; SVAP ⫽ sagittal vertical alignment–postural; CVA-G ⫽ coronal vertical alignment–
gait; SVA-C ⫽ sagittal vertical alignment– gait; APA-G ⫽ acromion-pelvic angle– gait; APA-G MAX ⫽ acromion-pelvic angle– gait maximum; APA-G MIN ⫽
acromion-pelvic angle– gait minimum; APA-G ROM ⫽ acromion-pelvic angle–
gait range of motion.
postures and body motions. Schizas et al evaluated gait
asymmetries in patients with idiopathic scoliosis using
vertical force measurements only.22 They utilized published gait data in normal subjects as a control group.8,15
In 20 of the 21 subjects, an asymmetry of at least one gait
parameter was noted. However, multiple regression
analysis showed that no association between the gait
asymmetry and curve direction, curve magnitude, or vertebral rotation existed.
There are several studies that have assessed the gait
patterns of patients after Harrington rod instrumentation and fusion.2– 4,6,10,16,25 These primarily investigated
the negative aspects of Harrington distraction instrumentation into the lumbar spine with subsequent loss of
lumbar lordosis and a creation of a flatback syndrome.6,26 This syndrome is associated with straightening of the lumbar spine, forward sagittal imbalance of
the entire spine on the pelvis, and flexion of the knees as
a compensatory mechanism to maintain as upright sagittal posture as possible. However, no study to date has
analyzed these patients objectively using gait and motion
analysis evaluations, although we have initiated this
project and are currently collecting patient data. Certainly, prevention of these flatback syndrome deformities
is currently the best means of treatment.
Results from our study demonstrate that lower extremity kinematics during gait are not changed 2 years
after spinal fusion surgery in patients with idiopathic
scoliosis. However, we did see a decrease in gait speed
secondary to a slower cadence and decreased stride
length in these patients after the fusion surgery. It is unclear whether this is due to restriction from the spinal
fusion placed on the trunk and torso or a possible deconditioning effect in these patients who potentially are less
active after their spinal fusion surgery than before it. The
body is not asked to perform at its maximum limits during gait and the surgical changes may have reduced the
limits, but not to gait limit thresholds.
Masso and Gorton were the first to evaluate standing
postural changes in scoliosis patients using optoelectronic measurements.20 They evaluated 33 subjects with
idiopathic scoliosis at approximately 13 months after
posterior instrumentation and fusion. They evaluated
these patients both in the coronal and transverse planes
and found consistent improvement in both planes. They
concluded that it was a useful, noninvasive technique to
evaluate patients with scoliosis and their deformity be-
E336 Spine • Volume 26 • Number 14 • 2001
fore and after surgery. Sagittal plane evaluation was not
included, nor were any gait data.
In this study sagittal spinal–pelvic parameters did not
show any differences after surgery on the radiographic,
standing postural, and gait analysis. This may be related
to the fact that none of these patients had a significant
preoperative sagittal malalignment to begin with, as is
the typical situation for patients with AIS treated with
segmental spinal instrumentation techniques.17 This is
not the case in patients treated with previous instrumentation systems (i.e., Harrington rods) that tended to remove lumbar lordosis and create long-term sagittal plane
problems.6,16 The most significant changes (improvements) were seen in the transverse plane where postoperative results demonstrated improved angular symmetry of the shoulders with respect to the pelvis at initial
contact.
It was interesting that no correlation existed between
the radiographic CVA and SVA variables and the postural and gait variables. This may reflect that the changes
before and after surgery being evaluated were small (⬍1
cm) in both planes. This could also be a function of
instructions and also the other differences that may occur
in standing posture when one is obtaining a standing
radiograph versus standing for a video electronic evaluation, as well as the position of the surface markers.
Vedantam et al have studied the importance of arm
alignment and positioning in obtaining a standing long
cassette spinal radiograph to evaluate sagittal balance.24
They found that positioning the arms at 30° was best
compared with 90° to the vertical. The most significant
difference was the increased negative sagittal vertical line
alignment as the arms were elevated from 30° to 90°.
Certainly, several limitations are present in this study,
probably the most significant being the lack of normal
controls. We elected to use the patient’s preoperative
data as their own control because our goal was to assess
the effects of spinal fusion surgery on these patients’ gait
and posture. There are several articles that have looked
at various gait parameters between healthy and scoliotic
patients previously.8,9 In addition, we would assume that
postural data for normal controls would be symmetric in
the coronal plane because of the lack of any deformity. In
the sagittal plane previous radiographic alignment parameters have shown similar sagittal balance values for
normal controls as was seen in the current study of scoliotic patients.24 In addition, there are certainly many
other confounding variables and technical procedures
performed on this study group that we were unable to
separate out from the scoliosis fusion surgery itself.
These included any effect that iliac crest bone grafting
would have on the pelvic parameters and separating the
effects that a formal open thoracotomy would have on
those patients that had an anterior spinal fusion. Also
included was the effect that prior bracing had on patients
who subsequently underwent a spinal fusion. Lastly,
there are several variables that we attempted to analyze,
such as differences in the lowest instrumented vertebrae,
and also the total number of fusion levels. However, in
our current study group there are not enough patients in
each one of these separate groups to make a viable statistical comparison. In addition, some of the measured
differences are small in magnitude, which makes clinical
interpretation of these differences to be analyzed with
caution.
In conclusion, patients with AIS undergoing a spinal
fusion show slight decreased gait speed at 2 years postoperation without any change in lower extremity kinematics. This decrease is due to a combination of a slower
cadence as well as a diminished stride length. Spinal and
pelvic balance parameters are improved in the coronal
plane radiographically, and also during standing posture
and gait, whereas sagittal balance parameters are unchanged radiographically and during standing posture
and gait. Lastly, significant improvements were seen in
the transverse plane where improved angular symmetry
of the shoulders with respect to the pelvis occurred during gait at the expense of a loss of transverse plane range
of motion. This type of analysis will further our understanding of the functional ramifications of performing
spinal fusion on these patients. Additional functional
analysis testing of trunk range of motion, trunk and abdominal strength, as well as aerobic analysis will also
complement the radiographic assessment to form a more
complete functional evaluation of these patients, providing important information for the patient, family, and
treating surgeon.
Key Points
● Lower extremity kinematics are unaffected by
spinal fusion in patients with adolescent idiopathic
scoliosis.
● Gait speed is decreased at 2 years postoperation
as a result of significantly reduced cadence and
stride length.
● Coronal spinal balance parameters are improved
radiographically and during standing posture and
gait.
● Sagittal balance parameters are unchanged on
radiographs and during standing posture and gait.
● Transverse plane parameters demonstrated improved angular symmetry during gait.
References
1. Aaro S, Dahlborn M. The effect of Harrington instrumentation on the longitudinal axis rotation of the apical vertebra and on the spinal and rib-cage
deformity in idiopathic scoliosis studied by computer tomography. Spine
1982;7:456 – 62.
2. Aaro S, Ohlen G. The effect of Harrington instrumentation on the sagittal
configuration and mobility of the spine in scoliosis. Spine 1983;8:570 –5.
3. Balderston RA, Winter RB, Moe JH, Bradford DS, Lonstein JE. Fusion to the
sacrum for non-paralytic scoliosis in the adult. Spine 1986;11:824 –9.
4. Boachie-Adjei O, Dendrinos GK, Ogilvie JW, et al. Fusion to the sacrum for
non-paralytic scoliosis in the adult. J Spinal Disord 1991;4:131– 41.
5. Burwell RG, Cole AA, Cook TA, et al. Pathogenesis of idiopathic scoliosis:
the Nottingham concept. Acta Orthop Belg 1992;58:33–58.
6. Cochran T, Irstam L, Nachemson A. Long-term anatomic and functional
Adolescent Idiopathic Scoliosis • Lenke et al E337
changes in patients with adolescent idiopathic scoliosis treated by Harrington rod fusion. Spine 1983;8:576 – 84.
7. Engsberg JR, Andrews JG. An analysis of the combined talocalcaneal/
talocrural joint during running support. Med Sci Sports Exerc 1987;19:275–
84.
8. Engsberg JR, Lee AG, Tedford KG, et al. Normative ground reaction force
data for walking. J Pediatr Orthop 1993;13:167–73.
9. Giakas G, Baltzopoulos V, Dangerfield PH, et al. Comparison of gait patterns between healthy and scoliotic patients using time and frequency domain analysis of ground reaction forces. Spine 1996;21:2235– 42.
10. Ginsburg HH, Goldstein L, Haake PW, et al. Longitudinal study of back
pain in postoperative idiopathic scoliosis: long-term follow-up, phase IV.
Presented at the Scoliosis Research Society 30th Annual Meeting, Asheville,
NC, 1995.
11. Gracovetsky S. A hypothesis for the role of the spine in human locomotion:
a challenge to current thinking. J Biomed Eng 1985;7:205–16.
12. Haher TR, Merola A, Zipnick RI, et al. Meta-analysis of surgical outcome in
adolescent idiopathic scoliosis: a 35-year English literature review of 11,000
patients. Spine 1995;20:1575– 84.
13. Hamill C, Lenke LG, Bridwell KH, et al. The use of pedicle screws to improve
correction in the lumbar spine of adolescent idiopathic scoliosis: is it warranted? Spine 1996;21:1241–9.
14. Hopf C, Scheidecker M, Steffan K, et al. Gait analysis in idiopathic scoliosis
before and after surgery: a comparison of pre- and postoperative muscle
activation pattern. Eur Spine J 1998;7:6 –11.
15. Khoo BC, Goh JC, Bose K. A biomechanical model to determine lumbosacral
loads during single stance phase in normal gait. Med Eng Phys 1995;17:27–
35.
16. LaGrone MO, Bradford DS, Moe JH, et al. Treatment of symptomatic flatback after spinal fusion. J Bone Joint Surg [Am] 1988;70:569 – 80.
17. Lenke LG, Bridwell KH, Baldus C, et al. Cotrel-Dubousset instrumentation
for adolescent idiopathic scoliosis. J Bone Joint Surg [Am] 1992;74:1056 –
67.
18. Lenke LG, Bridwell KH, Baldus C, et al. Ability of Cotrel-Dubousset instrumentation to preserve distal lumbar motion segments in adolescent idiopathic scoliosis. J Spinal Disord 1993;6:339 –50.
19. Lenke LG, Engsberg JR, Olree KS, et al. Trunk range of motion and gait
analysis following adolescent idiopathic scoliosis fusion. Poster presentation
at the Scoliosis Research Society Annual Meeting, St. Louis, MO, September
1997.
20. Masso PD, Gorton GE III. Quantifying changes in standing body segment
alignment following spinal instrumentation and fusion in idiopathic scoliosis
using an optoelectronic measurement system. Spine 2000;25:457– 62.
21. Richards BS, Birch JG, Hering JA, et al. Frontal plane and sagittal plane
balance following Cotrel-Dubousset instrumentation for idiopathic scoliosis. Spine 1989;17:733–7.
22. Schizas CG, Kramers-de Quervain IA, Stussi E, et al. Gait asymmetries in
patients with idiopathic scoliosis using vertical forces measurement only. Eur
Spine J 1998;7:95– 8.
23. Thurston AJ, Harris JD. Normal kinematics of the lumbar spine and pelvis.
Spine 1983;8:199 –205.
24. Vendantam R, Lenke LG, Keeney JA, et al. Comparison of standing sagittal
spinal alignment in asymptomatic adolescents and adults. Spine 1998;23:
211–5.
25. Vendantam R, Lenke LG, Linville D, et al. The effect on variation in arm
position on sagittal spinal alignment. Spine 2000;25:2204 –9.
26. Winter RB, Lonstein JE. Adult idiopathic scoliosis treated with Luque or
Harrington rods and sublaminar wiring. J Bone Joint Surg Am 1989;71:
1308 –13.
Address reprint requests to
Lawrence G. Lenke, MD
Department of Orthopaedic Surgery
Washington University School of Medicine
One Barnes-Jewish Plaza, Suite 11300
St. Louis, MO 63110
E-mail: lenkel@msnotes.wustl.edu