Share this article on:

Which Lateral Radiographic Positioning Technique Provides the Most Reliable and Functional Representation of a Patient’s Sagittal Balance?

Marks, Michelle PT, MA*; Stanford, Chriss MA*; Newton, Peter MD*†

doi: 10.1097/BRS.0b013e318199650a

Study Design. Prospective nonscoliotic cohort evaluation of the effects of various positions for obtaining standing lateral thoracolumbar radiographs.

Objective. The purpose of this study was to compare the effects of various upper extremity positions on thoracolumbar sagittal spinal alignment.

Summary of Background Data. The standing position used to capture a lateral plane radiograph can have marked effects on measurements of sagittal spinal alignment and may compound the variability between measurements from successive radiographs.

Methods. Twenty-two healthy female adolescents performed 3 repeated trials of 4 standing positions in a motion analysis laboratory. The positions included: (1) relaxed standing with arms at sides (CONTROL), (2) standing with fists overlying ipsilateral clavicles (CLAVICLE), (3) active shoulder flexion to 30° with elbows extended (30 ACTIVE), and (4) passive shoulder flexion to 30° with hand supports (30 PASSIVE).

Sagittal alignment of the spine was described by kyphosis, lordosis, and the sagittal vertical axis (SVA), all of which were measured from the positions of reflective markers attached to the surface of the back and pelvis. Differences between alignment measures obtained for each of the 3 radiographic positions relative to the functional position (CONTROL) were calculated. Mean differences were then compared between positions using repeated measures ANOVAs (α = 0.05).

Results. Relative to the CONTROL position, all other positions resulted in negative shifts in SVA (range = −1.1 cm–−4.6 cm), decreased kyphosis (range = −1°–−3°), and increased lordosis (4° for all positions). The shift in the SVA with the 30 PASSIVE position was significantly less than the other 2 positions (P < 0.05) and demonstrated the least variability.

Discussion. Standing with the hands supported while flexing the shoulders 30° during positioning for lateral spinopelvic radiographic acquisition resulted in an SVA and measures of sagittal plane curvature that were comparable with a functional standing position with arms at the side. This seems to be the best way to move the arms anterior to the spine with the least effect on overall sagittal balance.

In this prospective evaluation comparing the effects of variations in upper extremity positioning used for lateral spine radiographic acquisition, standing with hands supported in 30° of shoulder flexion resulted in an sagittal vertical axis and measures of sagittal plane curvature that are comparable with a functional standing position (with arms at sides) and demonstrated the least variability between repeated trials.

From the *Department of Orthopedic Surgery; Rady Children’s Hospital and Health Center, San Diego, CA; and †Department of Orthopedic Surgery, University of California San Diego, San Diego, CA.

Acknowledgment date: August 12, 2008. Revision date: October 23, 2008. Acceptance date: October 28, 2008.

The manuscript submitted does not contain information about medical device(s)/drug(s).

Corporate/Industry funds were received in support of this work. No benefits in any form have been or will be received from a commercial party related directly or indirectly to the subject of this manuscript.

Supported by a research grant from Medtronic-Sofamor Danek.

Address correspondence and reprint requests to Michelle Marks, PT, MA, 3030 Children’s Way, Suite 410, San Diego, CA 92123; E-mail:

The standing position used to capture a lateral thoracolumbar spinal radiograph can have marked effects on both the perceived alignment of the spine and the reproducibility of measurements between subsequent radiographs.

Assessment of the sagittal spinal alignment is becoming more common1–4 as outcomes of surgical treatment are increasingly recognized as dependent on the sagittal plane.5–9 The global sagittal balance of the spine is an important aspect for the clinician to consider in the evaluation and treatment of the adolescent and adult scoliotic spine. As scoliosis correction results in changes in both the coronal and sagittal planes, it is important to establish accurate and consistent methods for the assessment of a patient’s sagittal balance.

Ideally, the standing position during radiographic acquisition reflects a comfortable, functional, and naturally assumed posture. Unfortunately, a relaxed standing position (arms at sides) prevents adequate visualization of the spine on a lateral radiograph and flexion of the shoulders is required to bring the humerii anterior. Patients may therefore be instructed to adopt one of many possible standing positions, including varying degrees of shoulder flexion and knee flexion during lateral radiograph acquisition.10–17

Despite initial work investigating optimal positioning for lateral radiographic acquisition,15–18 variation of opinion still exists as to the “best” position. Flexion of the shoulders may be accomplished either actively (isometric contraction of the shoulder flexors) as in the traditional position of arms forward flexed 45° and elbows fully extended or passively, using a supporting platform or device for the hands or elbows. Both approaches have been used for acquiring radiographs, though it is not clear which produces spinal alignment measurements that best approximate those during a relaxed standing position.

Our previous work evaluating the sagittal vertical axis (SVA) as a measurement of global sagittal alignment of the spine via radiographs and 3-dimensional motion analysis revealed that radiographic positions involving active shoulder flexion resulted in negative shifts in the SVA (mean: −4.6) and do not produce sagittal spinal balance measures that are representative of a patient’s functional balance (mean SVA for relaxed standing: 0.9 cm). We concluded that further investigation of hand support was warranted.16 Vedantam et al recommended positioning the arms at 30° versus 90° of active shoulder flexion because it resulted in a less negative shift in the SVA.15

Horton et al evaluated 2 passive shoulder flexion positions (shoulder flexion of 90° and elbow extension vs. shoulder flexion of 60° and elbow flexion, both with gentle support provided through an IV pole with wheels).18 With regards to global sagittal balance, they found that in both positions, the SVA was shifted positive (the C7–S1 plumb line was anterior to the posterior margin of L5–S1 disc) and was significantly more positive (forward) in the 60° position than the 90° position.

An alternate method for visualization of the spine during lateral radiograph acquisition is the “fists on clavicle” position. In this position, patients are positioned with both fists on their ipsilateral clavicles, as an attempt to keep the humeri from blocking visualization of the spine or creating an additional force for the trunk to counterbalance as with the shoulders forward flexed. Horton et al found a positive shift in the SVA with this position (mean: 0.69) and recommend this position for its superior visualization of key vertebral landmarks when compared with other positions.18 Conversely, Faro et al found a negative shift in the SVA with this position but compared with a position of 45° of active shoulder flexion, the negative shift was less (−1.3 vs. −4.2, respectively).17

The fists on clavicle position was adopted at our institution as the standardized position for lateral radiograph acquisition; however, due to concerns regarding variations in patient positioning and variability in the SVA measurement, we felt that further investigation was warranted.

Therefore, the present study was conducted to compare the effects of 3 standing positions used to assess sagittal spinal alignment with 2 purposes in mind:

  1. To determine which position provides the most “functional representation” of a patient’s relaxed standing posture.
  2. To determine which position provides superior reliability across repeated trials.

The null hypothesis was that there is no significant difference between the positions.

Back to Top | Article Outline

Materials and Methods

Twenty-two nonscoliotic female adolescents between the ages of 12 and 20 years consented to participate in this prospective study. Institutional review board approval was granted before conducting any study procedures. Subjects completed an initial screening phase, including a medical history, physical examination, forward bend test, and observational gait analysis, to confirm absence of structural or functional abnormalities of the spine or pelvis. Subjects with a spinal abnormality or condition limiting normal movement were excluded from the study.

Back to Top | Article Outline

Subject Preparation

A physical therapist (author M.M.) palpated and identified all spinous processes and iliac spines by marking the surface of the skin of each subject using a washable ink pen. Three ½-inch diameter spherical, reflective markers were secured to the marks identifying the right and left anterior superior iliac spines (ASIS) and second sacral vertebrae (S2). One rigid cluster of 3 ¼-inch markers was attached to each pen mark overlying the vertebral spinous processes of T1, T5, T9, T12, and L3. The triads of markers were collectively used to identify the spatial orientation of the pelvis and 5 respective regions of thoracolumbar spine (Figure 1). Double-sided tape and adhesive spray were used to secure all markers.

Figure 1

Figure 1

Back to Top | Article Outline

Data Collection

An 8-camera infrared motion capture system (Motion Analysis Corp., Santa Rosa, CA) was used to record the 2-dimensional positions of the reflective markers at 60 Hz while each subject assumed 4 different standing positions (Figure 2): standing with arms resting on either side (CONTROL); standing with active shoulder flexion to 30° and elbows extended (30 ACTIVE); standing with passive shoulder flexion to 30° and elbows slightly flexed using “ski pole” type hand supports with rigid, stable bases (Figure 3) placed in front and to the side of each subject (30 PASSIVE); and standing with the elbows fully flexed and each fist placed over the ipsilateral clavicle (CLAVICLE). A standardized set of instructions for assuming each position were read to the subject, as indicated in Table 1.

Figure 2

Figure 2

Figure 3

Figure 3

Table 1

Table 1

Each position was held for several seconds and a total of 3 trials of each position were performed. A constant base of support (stance width) was assumed during each trial. Shoulder flexion to 30° for the ACTIVE and PASSIVE positions was verified using a goniometer. Subjects were completely repositioned between each trial.

Back to Top | Article Outline

Data Analysis

The 2-dimensional image coordinates of markers from each camera were reconstructed into 3-dimensional spatial coordinates using a centralized computer and a direct linear transformation. Marker trajectories for trials were interpolated using a first order/linear algorithm.

One triad of markers was used to define the pelvis and each spinal segment. The pelvis was defined using the ASIS and S2 markers and each marker cluster defined its corresponding spinal segment (as shown in Figure 4, ISB Standardization Committee, 2002). The 3 coordinates were used to create a coordinate system according to a standardized convention. The mediolateral axis of the pelvis was defined by the ASIS markers, with the anteroposterior axis passing through the mid point of the ASIS markers (in-plane with the S2 marker). The vertical axis of the pelvis was defined as the cross product of the former axes, creating an orthogonal coordinate system. Each spine segment was defined similarly, using the marker coordinates of the corresponding triad and an axis system as proposed by the International Society of Biomechanics.19

Figure 4

Figure 4

Sagittal balance and alignment of the spine were measured as the relative distances or angles in the principal (sagittal) plane. The SVA was calculated as the horizontal distance between a virtual coordinate representing the location of T1 and a virtual coordinate representing the location of S2. Kyphosis and lordosis were calculated as the relative angular position between the proximal and distal segments used to define the respective thoracic (T1 relative to T12) and lumbar (T12 relative to pelvis) regions. This validity of this technique has been previously reported by Leroux et al.2

Measurements of SVA, kyphosis, and lordosis were averaged over the 3 trials for each test position. Mean values for each test position were then normalized to the optimal position (CONTROL) by subtracting the CONTROL mean from each test position mean for each subject.

Normalized values for each test position were compared within subjects using repeated measures ANOVAs. One 2-tailed ANOVAs were used to compare differences in each variable (SVA, kyphosis, and lordosis) between the 3 test positions. In the presence of a significant main effect, Tukey post hoc tests were used to identify specific differences between test positions. All statistical tests were performed with Statistical Package for the Social Sciences (SPSS, Inc., Chicago, IL; version 12.0) using a level of significance of α = 0.05.

Back to Top | Article Outline


All 22 consented subjects met study entrance criteria. Subjects were aged 13 ± 2 years had a mean height and weight of 163 ± 7 cm and 55 ± 12 kg, respectively.

Figure 5 displays the mean normalized SVA values for each of the 3 experimental radiographic positions. The 0 position (y-axis) represents the mean value for CONTROL. All experimental positions resulted in a negative shift in SVA relative to CONTROL. The most profound shifts were observed for 30 ACTIVE (−4.6 ± 1.3 cm) and CLAVICLE (−3.7 ± 1.9 cm), both of which were significantly less than 30 PASSIVE (−1.1 ± 0.8 cm, P < 0.05). The standard deviation was smallest for the negative shift in the 30 PASSIVE position, indicating the least variability between trials (Figure 5).

Figure 5

Figure 5

Normalized means for kyphosis and lordosis are shown in Figures 6 and 7, respectively. Each test position resulted in slightly decreased kyphosis relative to CONTROL (0) (30 ACTIVE = −2° ± −7°, CLAVICLE = −3° ± −8°, 30 PASSIVE = −1° ± −6°), and increased lordosis (30 ACTIVE = 4° ± 7°, CLAVICLE = 4° ± 6°, 30 PASSIVE = 4° ± 5°) relative to CONTROL (0). The kyphosis change for the 30 PASSIVE position was the smallest, however, no differences were statistically significant (P > 0.05).

Figure 6

Figure 6

Figure 7

Figure 7

Back to Top | Article Outline


Standardized patient positioning for lateral radiographic acquisition is of critical importance as global sagittal balance and junctional sagittal alignment after surgical correction of adolescent idiopathic and adult scoliosis become increasingly important and more common outcome measures.5,20,21

This study was designed to evaluate 3 commonly used lateral radiograph patient positioning techniques. From our previous work, we found that active shoulder flexion results in a statistically significant, negative (posterior) shift (mean shift = −4.6 cm) in the sagittal vertical axis (SVA) and significant posterior tilting of the pelvis (mean change = −8°), relative to the relaxed standing position. Incorporation of knee flexion with active shoulder flexion resulted in an SVA that was more representative of relaxed standing when compared with active shoulder flexion alone, but greater intertrial variability in SVA was observed with this position. We concluded that, in cases where outcome is evaluated by comparing measurements on radiographs taken at different time points (e.g., pre- and postoperative measurements), active shoulder flexion without knee flexion is the most favorable position because differences between measurements of spinal alignment are less to be confounded by variations in body positioning.16

The results of the present study corroborate our previous findings in which all radiographic positions involving shoulder flexion resulted in a negative shift in SVA, relative to a functional standing position.

Standing with the hands supported, however, did result in a SVA that approached a functional standing position. Apparently, the supports provided stabilization through which subjects could assume a “near” neutral position. In this position, trunk extension (and a negative SVA shift) to counterbalance anterior displacement of the arms and maintain the body’s center of gravity over the base of support was not necessary. Between subjects variability for this position was smaller than for the other positions suggesting that the use of hand supports may provide a more predictable effect on SVA.

Where does this leave the fists on clavicle position? Our clinical experience with the fists on clavicle position instigated our current evaluation for 2 primary reasons: (1) we were experiencing a wide variety of interpretation of the fists on clavicle patient positioning by the numerous radiograph technicians at our institution and (2) negative shifts in the SVA were seen in consecutive radiograph images of patients who had not undergone treatment or a change in their deformity status. Additionally, our previous work had been retrospective in nature and had not directly compared the clavicle position to a position of shoulder flexion with the hands supported, which we felt was necessary to further determine the utility of this radiographic patient positioning technique. We previously had argued that the required use of an external supports in the radiology department was added equipment that would be a hindrance to the staff and patients, however, since implementing the radiograph positioning poles (which were created specifically for this purpose, Figure 3), we have experienced an improved compliance with the effort to standardize the patient positioning, and perhaps the external equipment reinforces the importance of the standardization compliance. The data from this current study indicating the least variability between trials in the position of standing with hands supported in 30° of shoulder flexion supports this claim. The use of these poles has been adopted by a multicenter study group for standardization of positioning for radiograph acquisition in multicenter studies.

Our current results do not agree with the findings of Horton et al 18 in which the authors compared 3 commonly used lateral radiograph positioning techniques: 90° shoulder flexion with arm support, 60° shoulder flexion with arm support, and the fists on clavicle position. The focus of their study was the evaluation of superiority in radiographic landmark visualization for which they found the clavicle position to be superior over the other 2 positions. However, with regards to sagittal alignment, they found that all 3 positions resulted in a positive shift in the SVA. These shift magnitudes were 0.37 cm, 1.35 cm, and 0.69 cm, respectively. These findings are in disagreement with previously published results in the negative shift in SVA with the fists on clavicle position17 and with our current findings (both with the fists on clavicle position and the shoulder flexion with arms supported position). The difference in these results may be explained by the patient population used in the studies, with an adult population used in the Horton et al study and adolescents used in the previously published results. Despite having similar regional and segmental sagittal alignments, adolescents had a significantly more negative SVA (mean: −5.6 cm) than adults (mean: −3.2 cm).14 In addition, the use of IV poles with wheels as the arm support in their study may have affected the patient’s ability to obtain a relaxed posture (which even though locked may subconsciously affect the patient’s stability) versus radiograph positioning poles with a stable base.

In our previous study, we proposed that the negative shift in SVA during active shoulder flexion was the result of extension of the entire spine. Results from our present study are generally in support of this premise. All positions on average were accompanied by a slight decrease in kyphosis and a slight increase in lordosis. Thus, the negative SVA shift is accomplished by extension of the entire spine relative to the pelvis, for these positions. Increased hip extension and posterior rotation of the pelvis may facilitate this shift.16

Back to Top | Article Outline


In summary, the use of hand supports with minimal shoulder flexion (to allow proper radiograph visualization of the upper thoracic spine) provides a standing position that produces sagittal spinal alignment measurements that best represent a patient’s relaxed standing position. This disproves our null hypothesis, in that there is a significantly smaller deviation in the negative shift in the SVA while standing with hands supported and flexing the shoulders 30° (30 passive) indicating that this position provides a more functional representation of a patient’s relaxed posture.

The use of hand supports may also provide improved standardization for patient positioning resulting in increased reliability of repeated measures. This is critically important when evaluating patient outcome after a treatment intervention.

Back to Top | Article Outline

Key Points

  • All radiographic positions involving shoulder flexion resulted in negative shifts of sagittal alignment.
  • The lateral radiograph position of standing with 30° of shoulder flexion and hands supported resulted in the least negative sagittal balance shift and exhibited the least variability.
  • The lateral radiograph position of standing with 30° of shoulder flexion and hands supported resulted in measures of sagittal plane curvature that were comparable with a functional standing position with arms at the side.
Back to Top | Article Outline


1. Slucky AV, Engelking B, Shufflebarger HL. A simple method for improving vertebral visualization on a scoliosis series lateral radiograph. Spine 1995;20:2570–1.
2. Leroux MA, Zabjek K, Simard G, et al. A noninvasive anthropometric technique for measuring Kyphosis and Lordosis: an application for scoliosis. Spine 2000;25:1690–94.
3. Leroux MA, Zabjek K, Simard G, et al. Estmated Kyphosis and Lordosis changes at follow-up in patients with idiopathic scoliosis. J Pediatr Orthop 2002;22:73–9.
4. Mac-Thiong JM, Labelle H, Berthonnaud E, et al. Sagittal spinopelvic balance in normal children and adolescents. Eur Spine J 2007;16:227–34.
5. Tanguay F, Mac-Thiong JM, de Guise JA, et al. Relation between the sagittal pelvic and lumbar spine geometries following surgical correction of adolescent idiopathic scoliosis. Eur Spine J 2006:19.
6. Smith JA, Deviren V, Berven S, et al. Does instrumented anterior scoliosis surgery lead to kyphosis, pseudarthrosis, or inadequate correction in adults? Spine 2002;27:529–34.
7. Ouellet JA, Johnston CE. Effect of grafting technique on the maintenance of coronal and sagittal correction in anterior treatment of scoliosis. Spine 2002;27:2129–35.
8. Betz RR, Harms J, Clements DH, et al. Comparison of anterior and posterior instrumentation for correction of adolescent thoracic idiopathic scoliosis. Spine 1999;24:2067–8.
9. De Jonge T, Dubousset JF, Illes T. Sagittal plane correction in idiopathic scoliosis. Spine 2002;27:754–60.
10. Bernhardt M, Bridwell KH. Segmental analysis of the sagittal plane alignment of the normal thoracic and lumbar spines and thoracolumbar junction. Spine 1989;14:717–21.
11. Gelb DE, Lenke LG Bridwell KH, et al. An analysis of sagittal spinal alignment in 100 asymptomatic middle and older aged volunteers. Spine 1995;20:1351–8.
12. Jackson RP, Kanemura T, Kawakami N, et al. Lumbopelvic lordosis and pelvic balance on repeated standing lateral radiographs of adult volunteers and untreated patients with constant low back pain. Spine 2000;25:575–86.
13. Jackson RP, McManus AC. Radiographic analysis of sagittal plane alignment and balance in standing volunteers and patients with low back pain matched for age, sex, and sixe: a prospective controlled clinical study. Spine 1994;19:1611–8.
14. Vedantum R, Lenke LG, Keeney JA, et al. Comparison of standing sagittal spinal alignment in asymptomatic adolescents and adults. Spine 1998;23:211–5.
15. Vedantam R, Lenke LG, Bridwell KH, et al. The effect of arm position on sagittal spinal alignment. Spine 2000;25:2204–9.
16. Marks MC, Stanford CF, Mahar AT, et al. Standing lateral radiographic positioning does not represent customary standing balance. Spine 2003;28:1176–82.
17. Faro FD, Marks MC, Pawelek J, et al. Evaluation of a functional position for lateral radiographic acquisition in adolescent idiopathic scoliosis. Spine 2004;29:2284–9.
18. Horton WC, Brown CW, Bridwell KH, et al. Is there an optimal patient stance for obtaining a lateral 36” radiograph? Spine 2005;30:427–33.
19. ISB Standardization and Terminology Committee. ISB recommendations on definitions of joint coordinate system of various joints for reporting of human joint motion–part I: ankle, hip and spine. J Biomech 2002;35:543–8.
20. Rhee JM, Bridwell KH, Won DS, et al. Sagittal plane analysis of adolescent idiopathic scoliosis: the effect of anterior versus posterior instrumentation. Spine 2002;27:2350–6.
21. Kim YJ, Lenke LG, Bridwell KH, et al. Proximal junctional kyphosis in adolescent idiopathic scoliosis after 3 different types of posterior segmental spinal instrumentation and fusions: incidence and risk factor analysis of 410 cases. Spine 2007;32:2731–8.

radiographic positioning; sagittal alignment; functional standing position; measurement reproducibility

© 2009 Lippincott Williams & Wilkins, Inc.