New Method for Predicting the Lumbar Lordosis Angle in Skeletal Material

Size: px

Start display at page:

Download "New Method for Predicting the Lumbar Lordosis Angle in Skeletal Material"

Flora Paul
5 years ago
Views:

1 THE ANATOMICAL RECORD 290: (2007) New Method for Predicting the Lumbar Lordosis Angle in Skeletal Material ELLA BEEN, 1,2 * HAYUTA PESSAH, 1 LAURENCE BEEN, 3 ARIE TAWIL, 4 AND SMADAR PELEG 1 1 Department of Anatomy and Anthropology, Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel 2 Department of Physical Therapy, Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel 3 Spinal Private Practice, Raanana, Israel 4 Spinal Rehabilitation Clinic, Raanana, Israel ABSTRACT Reconstructing the lordotic curvature of the lumbar spine in humans is essential for understanding their posture and locomotion. To date there is still no reliable method for predicting the lordotic curvature of disarticulated spines (in the absence of intervertebral disks). This article examines two possible methods for predicting the lordotic curvature of the lumbar spine. The first the traditional method is based on the degree of wedging of the vertebral bodies, and the second the suggested new method is based on a lateral view of the orientation of the inferior articular processes. We propose a linear regression model for predicting the lordotic curvature of the lumbar spine (lordosis angle) in disarticulated human spines. This model, derived directly from our new method, is a more reliable predictor of the lumbar lordosis angle in disarticulated spines. Anat Rec, 290: , Ó 2007 Wiley-Liss, Inc. Key words: articular process; vertebral spine; spinal curvature The evolutionary adaptation of the human spine to the upright position includes the development of the spinal curvatures, alterations in the size and wedge angle of the lumbar vertebral bodies, and changes in the shape and orientation of the articular processes (Putz, 1985; Latimer and Ward, 1993; Schendel et al., 1993; Boszczyk et al., 2001). Today, there is growing recognition of the functional and clinical importance of the spinal curvatures. These curvatures allow the loads applied to the spinal column to be efficiently absorbed. As such, they have a great bearing on the maintenance of upright posture and the efficacy of bipedal walking (Gracovetsky and Iacono, 1987; Farfan, 1995; Vaz et al., 2002; Vialle et al., 2005). To date, there is no valid method for calculating the lordotic curvature of the lumbar spine (lumbar lordosis) in skeletal material, nor is there any way to calculate the lordotic curvature of disarticulated spines such as those found at archeological sites. The method used for estimating the lordotic curvature in archeological material is based on the wedging of the vertebral bodies (Cunningham s index, or vertebral body wedge angle). This wedging results from the difference between the anterior and posterior heights of the vertebral body, measured from superior to inferior, and from its ventrodorsal length. The degree of wedging can only be used to estimate whether the lumbar spine is lordotic or kyphotic, but cannot be used to predict its degree of curvature (Cunningham, 1886; Digiovani et al., 1989; Scoles et al., 1991). The claim that vertebral body wedging relates to the lordotic curvature of the lumbar spine has found little support in the orthopedic literature. Most researchers measure the vertebral body and adjacent disk as one unit without considering the separate effect of vertebral *Correspondence to: Ella Been, Department of Anatomy and Anthropology, Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv 69978, Israel. Fax: beenella@ post.tau.ac.il Received 19 April 2007; Accepted 26 August 2007 DOI /ar Published online 31 October 2007 in Wiley InterScience (www. interscience.wiley.com). Ó 2007 WILEY-LISS, INC.

NEW METHOD FOR PREDICTING THE LUMBAR LORDOSIS ANGLE body wedging (Harrison et al., 2001; Kimura et al., 2001; Vaz et al., 2002; Roussouly et al., 2005; Vialle et al., 2005). However, Cheng et al.

2 NEW METHOD FOR PREDICTING THE LUMBAR LORDOSIS ANGLE body wedging (Harrison et al., 2001; Kimura et al., 2001; Vaz et al., 2002; Roussouly et al., 2005; Vialle et al., 2005). However, Cheng et al. (1998) found moderate to weak correlations between vertebral body wedging (Cunningham s index) and the lumbar lordosis angle. The anterior pillar of the vertebral spine is composed of the vertebral bodies and intervertebral disks, and it is this anterior pillar that determines the lordotic curvature of the lumbar spine (Louis, 1985). It is reasonable to assume that the morphology of the posterior pillar, which is formed by the superior and inferior articular processes and the lamina, is influenced by the curvature exhibited by the lumbar spine. Our inspection of the inferior articular process of humans on the sagittal plane has revealed the correlation of that orientation with lordotic curvature (lumbar lordosis). To the best of our knowledge, this is the first time the interaction between the lordotic curvature of the lumbar spine and the orientation of the inferior articular process on the sagittal plane has been investigated. Such a correlation, if it exists, could be used as a reference for predicting the degree of lordotic curvature in disarticulated lumbar spines (i.e., in the absence of intervertebral disks). In this study, we propose a new method for predicting the lordotic curvatures of disarticulated spines. The goal of this study is threefold: first, to test the relationship between the lordotic curvature of the lumbar spine and vertebral body wedging; second, to test the relationship between the lordotic curvature and the inferior articular process angle; and third, to provide the best tool for predicting the lordotic curvature using only bony landmarks within the lumbar spine. Fig Lumbar lordosis angle (L,S) based on the cobb method. MATERIALS AND METHODS Lateral radiographs of the lumbar spine of 379 subjects (patients at spinal clinics, Raanana, Israel) were examined. All radiographs were part of a routine evaluation performed in the spinal clinics to assess posture and detect abnormality in the spinal column. (Patients attend the spinal clinics for a variety of musculoskeletal complaints, including, for example, neck and back pain, headaches, or numbness of the hand or arm.) Of the 379 radiographs, 106 (56 men and 50 women) were selected according to the following criteria: adult subjects ( years of age, from 20 to 50 years old) with no history of spinal surgery and no radiographic abnormality (degeneration, scoliosis, spondylolysis/listhesis, reduced disk height, etc.) detected. In all cases, standard lateral lumbar radiographs were obtained with subjects standing in a comfortable position with knees straight and arms folded on the chest. All subjects signed an informed consent form, and the research was approved by the Helsinki committee of Tel-Aviv University. For each of the five lumbar vertebrae, three lines were drawn (Figs. 1, 2): along the superior endplate of the vertebral body (also on the first sacral vertebra); along the inferior endplate of the vertebral body; and along the anterior border of the inferior articular process. These lines were used to measure four angles: the lordosis angle (L 1 S 1 ) between the superior endplate of L 1 and the superior endplate of S 1 ; the body wedge angle (B) between the superior and inferior endplates of a single vertebra; the total segmental angle (T) between the superior endplate of one vertebra and the superior endplate of the successive vertebra; and the inferior articular process angle (AP) between the superior endplate and the anterior border of the inferior articular process of the same vertebra. Measurements B, T, and AP were taken for each of the five lumbar segments. The lordosis angle is shown in Figure 1, and the other three angles in Figure 2. All measurements were taken by the same investigator (E.B.) using a 25-cm Jamar goniometer with a 3608 scale in 18 increments. Measurements B, T, and AP were used to calculate ST, the sum of the lumbar L 1 L 5 segmental angles; SB, the sum of the lumbar L 1 L 5 body wedge angles; and SAP, the sum of the lumbar L 1 L 5 inferior articular process angles. Intraobserver reproducibility was assessed on the basis of five radiographs, remeasured by EB (reproducibility test: Intraobserver reproducibility was found to be high for the L1S1, B, and T angles [>99.7%, P < 0.001], and for the AP angles (<88.3%, P <.001). Kolmogorov-Smirnov tests were carried out to ensure normal distribution; unpaired t-tests were used to identify differences between men and women; correlation (Pearson) analysis was performed to investigate the relationship between all of the morphological parameters (L 1 S 1 ; ST; SB; SAP;T,B,andAPofeachspinal segment). Linear regression models based on these correlations were developed for the lumbar lordosis angle and for the segmental angles. The significance value was determined at 0.05.

3 1570 BEEN ET AL. Fig. 2. Segmental measurements. L3B, vertebral body angle of the third lumbar vertebra; L3T, total segmental angle of the third lumbar vertebra; L4AP, inferior articular process angle of the fourth lumbar vertebra. value of the lumbar spine (8) Vertebra TABLE 1. Lumbar spinal angles a value of L 1 (8) value of L 2 (8) value of L 3 (8) value of L 4 (8) value of L 5 (8) L 1 S Segment (T) SB Body (B) SAP AP a Positive values indicate lordotic wedging; negative values indicate kyphotic wedging. L 1 S 1, lordosis angle; SB, the sum of body wedge angles; SAP, the sum of inferior articular process angles; St, the sum of the lumbar segmental angles; B, body wedge angle; AP, inferior articular process angle; T, segmental angle. RESULTS The average wedging of the vertebral bodies is kyphotic in the upper lumbar vertebrae (e.g., for L 1 ), while there is no wedging in L 3 ( ), and lordotic wedging in the lower lumbar vertebrae (e.g., for L 5 ). The angular values of the AP gradually increase along the lumbar spine, from 948 at L 1 to 1138 at L 5. The gradual increase of the AP angulations is similar to the gradual change in T, which increases from 38 at L 1 to 198 at L 5 (Table 1). All angles were normally distributed, with no significant differences between men and women (P < 0.05). Tables 2 and 3 summarize the major correlations between the lumbar spinal angles. As expected, the lumbar lordosis angle (L 1 S 1 ) is highly correlated with the sum of the total segmental angles (ST; r 5.950; P < 0.01). High correlation is also found between the lumbar lordosis angle and the sum of the AP angles (SAP; r 5.787; P < 0.01; Fig. 3), and between the lumbar lordosis angle and the L 2 segmental angle (r 5.732; P < 0.01). Surprisingly, the angle of the inferior articular process of the second lumbar vertebra (L 2 AP) is moderately to highly correlated with the lumbar lordosis angle (L 1 S 1 ; r 5.592; P < 0.01). The sum of body wedge angles (SB) and the lumbar lordosis angle moderately correlate (r 5.464, P < 0.01; Fig. 4).

4 NEW METHOD FOR PREDICTING THE LUMBAR LORDOSIS ANGLE 1571 TABLE 2. Correlation (Pearson) matrix for the main spinal angles L 1 S 1 SB SAP ST L 1 T L 2 T L 2 AP L 3 T L 4 T L 5 T L 1 S SB ** SAP **0.787 ** ST **0.950 **0.451 ** L 1 T ** **0.291 ** L 2 T **0.732 **0.361 **0.623 **0.732 ** L 2 AP **0.592 **0.294 **0.685 **0.515 **0.264 ** L 3 T **0.678 **0.384 **0.570 ** **0.646 * L 4 T **0.600 **0.413 **0.403 ** **0.399 *0.251 ** L 5 T ** *0.197 ** a For an explanation of the abbreviations, see Table 1. *P < **P < TABLE 3. Correlation between specific measurements of each spinal segment B AP L 1 T **0.449 **0.373 L 2 T **0.567 **0.588 L 3 T **0.457 **0.565 L 4 T **0.549 **0.493 L 5 T **0.323 **0.513 a For an explanation of the abbreviations, see Table 1. **P < Fig. 4. Correlation between SB and the lumbar lordosis angle (L,S); r Fig. 3. Correlation between SAP and the lumbar lordosis angle (L,S); r The three segments in the middle of the lumbar spine (L 2,L 3, and L 4 ) are moderately correlated (.399 r.646) and moderately to highly correlated with the lumbar lordosis angle (.600 r.732). The two segments at the ends of the lumbar spine (L 1,L 5 )arenotcorrelated with any lumbar segments or with the lumbar lordosis angle (Table 2). Table 3 reveals that, at each spinal segment, the body wedge angle and the articular process angle correlate only moderately with the total segmental angle. We calculated three linear regression models for predicting the lumbar lordosis angle, each based on one of the following: SAP, L 2 AP, and SB (Table 4). The model based on the orientation of the inferior articular processes (SAP) explains 62% of the variability found in the lumbar lordosis angle (R ). In contrast, the model based on L 2 AP explains only 35% of this variability (R ), and the model based on the sum of the vertebral body wedge angles explains only 21% (R ). The model based on the orientation of the inferior articular processes also presents the lowest standard

5 1572 BEEN ET AL. TABLE 4. Simple linear correlations between the main spinal parameters Linear regression r R 2 SE Lordosis angle (L 1 S 1 ) predicted from the L 1 S SAP sum of AP angles Lordosis angle (L 1 S 1 ) predicted from L 2 AP L 1 S L 2 AP Lordosis angle (L 1 S 1 ) predicted from the L 1 S SB sum of body angles L 1 segment angle predicted from L 1 B L 1 T L 1 B L 2 segment angle predicted from L 2 B L 2 T L 2 B L 2 segment angle predicted from L 2 AP L 2 T L 2 AP L 3 segment angle predicted from L 3 B L 3 T L 3 B L 3 segment angle predicted from L 3 AP L 3 T L 3 AP L 4 segment angle predicted from L 4 B L 4 T L 4 B L 4 segment angle predicted from L 4 AP L 4 T L 4 AP L 5 segment angle predicted from L 5 AP L 5 T L 5 AP a For an explanation of the abbreviations, see Table 1. error (S E 5 6.8). The standard error for the two other models (based on L 2 AP and on SB) are S E and S E 5 9.6, respectively. The models for the segmental angles (from L 1 to L 5 ) can only explain 20 34% of the variability found in the segmental angles of each spinal level (0.201 R 2.345). DISCUSSION This study provides a new method for predicting the lumbar lordosis angle of disarticulated spines using bony landmarks within the spine. It also yields basic data for several angular measurements of the lumbar spine and the relationship between them. Many researchers have measured the angles of vertebral segments and their contribution to the lumbar lordosis angle in living human subjects (Harrison et al., 2001; Kimura et al., 2001; Vaz et al., 2002; Roussouly et al., 2005; Vialle et al., 2005), but only Cheng et al. (1998) have ever investigated the relationship between vertebrae morphology and the degree of lumbar lordosis. Understanding this relationship is essential for predicting lordosis in disarticulated spines, such as those found at archeological sites. The relationship between the lumbar lordosis angle and the orientation of the inferior articular process as seen in a lateral view is crucial for predicting the degree of lordosis in disarticulated lumbar spines. We have shown that the degree of lumbar lordosis (L 1 S 1 ) and the cumulative degree of orientation of the inferior articular processes (SAP) are closely correlated. We have also shown that the best linear regression model for predicting the lumbar lordosis angle is the one based on the orientation of the inferior articular process (SAP). This model explains 62% of the variability found in the lumbar lordosis angle in the studied population (R ) and has the lowest standard error (S E 5 6.8). Therefore, we suggest that, when the lumbar spine is disarticulated, the angle of lordosis can be predicted most reliably by the linear regression model that is based on inferior articular process orientation (Table 4). The second best model for predicting the lumbar lordosis angle is the one based on the orientation of the inferior articular process of the second lumbar vertebra (R , S E 5 8.9). That the L 2 segmental angle is highly correlated with the lumbar lordosis angle might explain the high correlation between L 2 AP and the degree of lumbar lordosis. Because vertebral body wedge angles (SB) only moderately correlate with the lumbar lordosis angle, the model that is based on this correlation should be used only when the inferior articular processes are missing (R ; S E 5 9.6). At each lumbar spinal level, the segmental angle, body wedge angle, and inferior articular process angle are only moderately correlated. Because the correlations are not high, our predictions of the curvatures of individual segments are less reliable than our predictions of the degree of lumbar lordosis (L 1 S 1 ). Consequently, we suggest that the most reliable prediction for the curvature of the L 1 segment can be obtained using the linear regression model based on vertebral body wedging; for the L 2 L 4 segments, the vertebral body wedge angle and the inferior articular process angle can be used. For the L 5 segment, the most reliable calculation is the one based on the inferior articular process angle. In practice, we suggest that, for disarticulated spines, this method be used by performing a lateral x-ray of the lumbar vertebrae, drawing the lines, measuring the angles accordingly, and then predicting the lordosis angle. It should be borne in mind that the derived values are valid for humans only and may not be reliable for all ethnic groups or for other species. To apply this method for other species (fossils included), one would need to demonstrate lower values of the inferior articular process angle (400 < SAP < 440) in the lumbar spines of primates that lack lordosis. Our lumbar lordosis angle values ( ) fall within the range presented in the orthopedic literature, for example, in Harrison et al. (2001; ), Vialle et al. (2005; ), and Lin et al. (1992; ). Therefore, we believe our results to be representative of the population at hand. The study is limited by a lack of information regarding the weight and activity level of the subjects. ACKNOWLEDGMENT The authors thank Yoel Rak for his comments. LITERATURE CITED Boszczyk BM, Boszczyk AA, Putz R Comparative and functional anatomy of the mammalian lumbar spine. Anat Rec 264:

6 NEW METHOD FOR PREDICTING THE LUMBAR LORDOSIS ANGLE Cheng XG, Sun Y, Boonen S, Nicholson P, Brys P, Dequeker J, Felsenberg D Measurement of vertebral shape by radiographic morphometry: sex differences and relationship with vertebral level and lumbar lordosis. Skeletal Radiol 27: Cunningham DJ The lumbar curve in man and apes. Nature 33: Digiovanni BF, Scoles PV, Latimer BM Anterior extension of the thoracic vertebral bodies in Scheuermann s kyphosis. Spine 14: Farfan HF Form and function of the musculoskeletal system as revealed by mathematical analysis of the lumbar spine. Spine 20: Gracovetsky S, Iacono S Energy transfer in the spinal cord. J Biomed Eng 9: Harrison DE, Harrison DD, Cailliet R, Janik TJ, Holland B Radiographic analysis of lumbar lordosis. Spine 26:E235 E242. Kimura S, Steinbach GC, Watenpaugh DE, Hargens AR Lumbar spine disc height and curvature responses to an axial load generated by a compression device compatible with magnetic resonance imaging. Spine 26: Latimer B, Ward CV The thoracic and lumbar vertebrae. In: Leakey RE, Walker A, editors. The Nariokotome Homo erectus skeleton. Cambridge MA: Harvard University. p Lin RM, Jo IM, Yu CY Lumbar lordosis: normal adults. J Formos Med Assoc 91: Louis R Spinal stability as defined by the three-column spine concept. Anat Clin 7:33. Putz R The functional morphology of the superior articular process of the lumbar vertebrae. J Anat 143: Roussouly P, Gollogy S, Berthhonnaud E, Dimnet J Classification of normal variation in the sagittal alignment of the human lumbar spine and pelvis in the standing position. Spine 30: Schendel M, Wood KB, Butterman GR, Lewis JL, Ogilvie JW Experimental measurement of ligament force, facet force, and segment motion in the human lumbar spine. J Biomech 26: Scoles PV, Latimer BM, Digiovanni BF, Vargo E, Bauza S, Jellema LM Vertebral alterations in Scheuermann s kyphosis. Spine 16: Vaz G, Roussouly P, Berthhonnaud E, Dimnet J Sagittal morphology and equilibrium of pelvis and spine. Eur Spine J 11: Vialle R, Levassor N, Rillardon L, Templier A, Skalli W, Guigui P Radiographic analysis of the sagittal alignment and balance of the spine in asymptomatic subjects. J Bone Joint Surg Am 87:

River North Pain Management Consultants, S.C., Axel Vargas, M.D., Regional Anesthesiology and Interventional Pain Management.

River North Pain Management Consultants, S.C., Axel Vargas, M.D., Regional Anesthesiology and Interventional Pain Management. Chicago, Illinois, 60611 Phone: (888) 951-6471 Fax: (888) 961-6471 Clinical