When a person stands on two feet, the total posture must be well balanced. If the position of the skeleton changes at one site, other sites must compensate to restore the balance (1). The consequences of frontal, axial, and sagittal imbalance, which occurs in persons with scoliosis, are well known. Any imbalance is energetically taxing (2), and thus the body must find a compensatory mechanism to reestablish balance. Such compensatory mechanisms lead to muscle contracture and fatigue, which cause pain. For these reasons, examinations of sagittal imbalance are frequently performed.
The sagittal shape of the spine plays an important role in the development of many spinal disorders (1,3,4). Well known is the close relation between the sagittal shape of the spine and low back pain syndrome (4–6). Although the relation between the gravity line of the body and spinal parameters (T9 tilting, lordosis, kyphosis) has been investigated frequently (7–9), the relation between the spinal and pelvic parameters has not been explored thoroughly (6,10–12).
The primary aims of our multicenter study were to evaluate the sagittal shape of the spine and the position of the pelvis in the space in isthmic spondylolisthesis and to explore the compensatory mechanisms by which patients tend to regain their balance. In addition, we investigated the importance of two new sagittal parameters, the overall inclination of the spine, and the amplitude of the curvatures from a biomechanical perspective.
After physical examination, radiographic examinations are used most frequently to diagnose spinal disorders, both in clinical practice and research. However, the evaluation of radiographs, particularly when several variables are involved, is often laborious, so the development of a means to accelerate and simplify these measurements would be valuable. Such methods are based on the digitization of radiographs and computerization of the analysis. New software was developed in 1998 for computer-assisted measurements of sagittal plane radiographs, with a preliminary evaluation for postural radiographs (13).
MATERIALS AND METHODS
We studied 30 lateral radiographs of the entire spine of 30 healthy persons. Table 1 provides general demographics of the volunteers and the affected patients. Inclusion criteria for the adult volunteers were age (range, 30 to 39 years), no previous spinal surgery, no spinal disease, and no history of hip and back pain or problems. After they provided informed consent, these volunteers were examined clinically by two independent orthopedic surgeons and underwent radiographic examinations to establish the normative physiologic values.
The radiographs were taken in the radiology department of the Mulhouse Hospital CHU, in Garches, France. For all 30 volunteers, left-to-right standing 30 × 90 cm lateral and anteroposterior radiographs of the pelvis and the entire spine were taken by the same technologist, using the same X-ray machine, with a fixed cassette-to-X-ray-source distance of 185 cm. A requirement for the standing radiographs was the visibility of the femoral heads of both hips. The volunteer's right side was situated against the cassette. Each volunteer was positioned and asked to stand straight, but relaxed. The knees were extended as much as possible, with the hips perpendicular to the film, and the arms were held out slightly below the chest level and supported by the back of the same chair.
We examined 48 patients (32 men and 16 women) with isthmic spondylolisthesis. The inclusion criteria for these patients were no previous spinal surgery and no degenerative changes on their spines. These patients were examined and treated in four French orthopedic centers: the Choisy Hospital (Paris), the Tripode Hospital (Bordeaux), the Hotel Dieu Hospital (Nantes), and the Pitié Hospital (Paris). The patients were examined clinically by two independent orthopedic surgeons and radiographs were taken before operation.
All patients signed consent forms allowing their clinical data to be used for research. In all these centers, the radiologic protocol was the same, as described for the healthy volunteer group. All radiographs were sent to the Laboratory of Biomechanics of ENSAM (Paris) with a completed checklist for each participant.
After scanning the lateral radiographs (Vidar-12 radiograph scanner [Vidar System Corp., Herndon, VA, U.S.A.]; 75 dpi resolution, 12-bit gray scale), we used SpineView software (SurgiView, Paris, France) to analyze the 13 independent variables listed here. The validation protocol of the software measurements was described previously in this journal (12).
We divided the measured variables into two groups: pelvic and spinal parameters (Fig. 1). We defined the pelvic morphologic and positional parameters according to the Duval-Beaupére criteria (14). The morphologic parameters include the sacrofemoral anatomic constant (SFAC or incidence) and thickness of the pelvis, and the positional parameters are the sacral slope, sacrofemoral tilting, and overhang.
Most of the spinal parameters have been described (3,5,9,14–16) and are widely used (e.g., lumbar L1-L5 lordosis, T9 tilting angle, T9 projection, thoracic T4-T12 kyphosis, and the lumbar angle). The software calculates lordosis and kyphosis as the angles of intersection of the lines perpendicular to the anterior wall of the corresponding vertebrae.
We considered two additional parameters (13): the overall inclination of the spine, which is the angle between the vertical and the best-fitted straight line (in a mean square sense) considering the anterior wall of the vertebrae from L5 to C2; and the amplitude of the spinal curvatures, which is the maximum distance of the vertebral anterior walls from the straight line best fitted to the spine in kyphosis and lordosis, considering the direction defined in connection with the inclination of the spine (Fig. 1). The program calculates angles in degrees, and all distances are expressed as percentages of the length of the sacral plate to avoid the effect of X-ray beam divergence.
We used commercial statistical software (StatView 4.5) (17) to analyze differences between normal and isthmic spondylolisthesis results. In the case of normal distribution, we used the Student t test to analyze the differences between groups, and we used the traditional Pearson correlation test to evaluate the correlation between all variables. To determine the significance of the correlation coefficients, we used the Fisher z-test. Statistical significance was set at a probability level of 5% or less.
The radiographs of the 30 healthy volunteers were labeled normal by two independent orthopedic surgeons, because the anteroposterior and lateral projections revealed no spinal disorders such as spondylolysis, spondylolisthesis, scoliosis, apophyseal osteochondrosis, Scheuermann kyphosis, or degenerative disk changes.
An analysis of the checklist of the patients (Table 1) revealed that most patients in the isthmic spondylolisthesis group had participated in some sporting activity before their symptoms occurred. The body mass index (i.e., body weight expressed in kilograms per height in square meters) did not indicate any significant difference between the groups at a probability level of 5% or less.
The Meyerding classification (18) showed that 5% of the patients in the spondylolisthesis group had spondylolysis without slipping, whereas 31% had grade I, 58% had grade II, and 6% had grade III spondylolisthesis.
Intergroup Comparison of Measured Variables
The SFAC, the sacral slope, and the degree of L1–L5 lordosis were greater in the patients with isthmic spondylolisthesis than in the healthy, asymptomatic volunteers. The values of sacrofemoral tilting, overhang, T9 projection, and T9 tilting were greater in the patients than in the healthy volunteers. We found no significant difference between the groups with regard to the thickness of the pelvis, the lumbar angle, the degree of T4–T12 kyphosis, the sagittal tilting angle, the amplitude of curvatures, or the inclination of the spine (Figs. 2 to 4).
Correlations Between Measured Variables
Table 2 lists the significant correlations between the various pelvic and spinal parameters. The numeric value is the correlation coefficient, whereas the asterisks denote the strength of the significance (*0.01 < p < 0.05; **0.001 < p < 0.01; and ***p < 0.001).
According to studies by Bernhardt and Bridwell (3) and Legaye et al. (14), the center of gravity is located approximately at the level of the T9 vertebra. The body's line of gravity passes through many vertebrae and leaves the spine at the promontory (19). The sacrum is a double-lever arm, which is supported by the sacroiliac joint, where the forces pass through the spine to the pelvis and legs. The posterior arm of the sacrum is fixed to the spine and the pelvis by the sacroiliac and lumbosacral ligaments.
The high value of SFAC and the sacral slope show that the sacrum is positioned more horizontally in the patients with spondylolisthesis than in the healthy volunteers. A horizontal sacrum is characteristic of quadrupeds, and this upright position implies maladaptation in humans (2). The horizontally positioned sacrum and hyperlordosis cause the shearing component of gravity to be greater than the compressive force (Fig. 5). The increased shearing force causes fracture of the dysplastic interarticular part of the vertebra. The high degree of the SFAC and sacrofemoral tilting shows an alteration in the spatial relation between the hips and the sacropelvis. The two acetabula are located well anterior to the lumbosacral junction, where the gravity line leaves the spine. This instability leads to spondylolisthesis: The L5 vertebral body slips anteriorly to regain balance by maintaining the gravity line above the hips (20). The higher value of the T9 tilting shows that the gravity line approached the femoral heads and allowed the patients to assume a new balance by virtue of the slipping. In this condition, the correlation matrix of the patients was similar to that of the healthy volunteers', which suggests that slipping of the vertebral body facilitated such balance, which we observed in our asymptomatic volunteers.
The values of the two new variables did not differ significantly between the asymptomatic and isthmic spondylolisthesis groups. Further investigation is needed to judge their association with age or other disorders (such as disk herniation).
A hereditary factor, the SFAC determines the sagittal curvatures through the sacral slope in healthy persons (10). If this relation is missing, problems develop because of the physical laws just described. In our study, the SFAC correlated well with the degree of slipping.
To achieve excellent results and avoid any sagittal imbalances, the sagittal balance of a patient must be determine before surgery. Computerized analysis permits data registration in a multicenter context and provides the opportunity for telediagnosis and consultation with other specialists.
The authors thank Drs. Christian Mazel (Department Orthopedic Surgery, Choisy Hospital, Paris, France), Jean-Marc Vital (Department of Orthopedic Surgery, Tripode Hospital, Bordeaux, France), Joel Delecrin (Department of Orthopedic Surgery, Hotel Dieu Hospital, Nantes, France), and Gerard Saillant (Department of Orthopedic Surgery, Pitié Hospital, Paris, France), for providing the study patients and volunteers, and Claude Kauffman (Laboratory of Biomechanics of ENSAM, Paris, France) for his assistance in the development of the SpineView software.
1. Farfan HF, Huberdeau RM, Dubow HI. Lumbar intervertebral disc degeneration. J Bone Joint Surg 1972; 54A:492—509.
2. Gracovetsky S, Farfan H. The optimum spine. Spine 1986; 11:543–73.
3. Bernhardt M, Bridwell KH. Segmental analysis of the sagittal plane alignment of the normal thoracic and lumbar spines and thoracolumbar junction. Spine 1989; 4:717–21.
4. Eiji I. Roentgenographic analysis of posture in spinal osteoporotics. Spine 1991; 16:750–6.
5. Jackson PR, McManus CA. Radiographic analysis of sagittal plane alignment and balance in standing volunteers and patients with low back pain matched for age, sex and size. Spine 1994; 19:1611–18.
6. Marty C, Legaye J, Duval-Beaupere G. Equilibre rachidien sagittal normal; ses relations avec les parameters pelviens, ses disfonctionnements: Cause de lombalgies. Resonances Europeennes du Rachis 1997; 15:21–8.
7. Gelb ED, Lenke GL, Bridwell KH, et al. An analysis of sagittal spinal alignment in 100 asymptomatic middle and older aged volunteers. Spine 1995; 20:1351–8.
8. Vedantam R, Lenke GL, Keeney JA, et al. Comparison of standing sagittal spinal alignment in asymptomatic adolescents and adults. Spine 1998; 23:211–15.
9. Jackson PR, Peterson DM, McManus CA, et al. Compensatory spinopelvic balance over the hip axis and better reliability in measuring lordosis to the pelvic radius on standing lateral radiographs of adult volunteers and patients. Spine 1998; 23:1750–67.
10. Boisaubert B, Montigny JP, Duval-Beaupére G, et al. Incidence, sacrum and spondylolisthesis. Rachis 1997; 9:187–92.
11. Kauffman C, Templier A, Pomero V, et al. A software for the quantitative analysis of static and dynamic radiographs of the lumbar spine. Third International ARGOS Meeting, Paris, France, January 29–31, 1999.
12. Rajnics P, Pomero V, Illes T, et al. Computer-assisted assessment of spinal sagittal plane radiographs. J Spinal Disord 2001; 14:135–42.
13. Templier A. Parameters et methodes d'evaluation des implantes rachidiens lombaires. PhD. thesis. Laboratoire de Biomechanique de l'ENSAM, 1998.
14. Legaye J, Hecquet J, Marty C, et al. Sagittal equilibration of the spine. Rachis 1993; 5:215–26.
15. Stagnara P, De Mauroy JC, Dran G, et al. Reciprocal angulation of vertebral bodies in a sagittal plane: Approach to references in the evaluation of kyphosis and lordosis. Spine 1982; 7:335–42.
16. Fernand R, Fox ED. Evaluation of lumbar lordosis. Spine 1985; 10:799–803.
17. Statview 4.5. Berkeley, CA: Abacus Concepts, 1996.
18. Meyerding HW. Spondylolisthesis. Surg Gynecol Obstet 1932; 54:371–7.
19. Szentágothai J, Kiss J. Functional Anatomy
. 6th Ed. Budapest, Hungary: Semmelweis, 1997.
20. Schwab FJ, Farcy JPC, Roye DP. The sagittal pelvic tilt index as a criterion in the evaluation of spondylolisthesis. Spine 1997; 22:1661–7.