Clinical Biomechanics 17 (2002) 361–367 www.elsevier.com/locate/clinbiomech
Influence of age on cervicothoracic spinal curvature: An ex vivo radiographic survey Jeffrey J.W. Boyle a, Nicholas Milne b, Kevin P. Singer
a,c,*
a
Department of Surgery, Centre for Musculoskeletal Studies, The University of Western Australia, Royal Perth Hospital, Level 2, MRF Building, WA Medical Res. Inst. Bldg., Rear 50 Murray Street, Perth, WA 6000, Australia Department of Anatomy and Human Biology, Centre for Musculoskeletal Studies, The University of Western Australia, Royal Perth Hospital, Level 2, MRF Building, WA Medical Res. Inst. Bldg., Rear 50 Murray Street, Perth, WA 6000, Australia c Department of Neuropathology, Centre for Musculoskeletal Studies, The University of Western Australia, Royal Perth Hospital, Level 2, MRF Building, WA Medical Res. Inst. Bldg., Rear 50 Murray Street, Perth, WA 6000, Australia
b
Received 26 November 2001; accepted 17 April 2002
Abstract Objective. To define the post-mortem cervicothoracic spinal curvature relative to age. Design. Spinal curvature assessment of lateral cervicothoracic radiographs. Background. A late consequence of age is the progressive accentuation of spinal curvatures, particularly the thoracic kyphosis. Little is known about the influence of the kyphosis on the alignment of the cervical spine. Method. One hundred and seventy two lateral spinal radiographs (113 males, 59 females) were analysed using two procedures: (1) sagittal curve deformation angles were derived, according to the method of Cobb, for thoracic (T1–T12), cervical (C2–C7) and cervicothoracic junctional regions (C6–T4); and (2) the cervicothoracic curvatures were digitised (C2–T12), to derive the apex of both curves and the inflexion point. Results. A significantly increasing thoracic spinal curvature was determined for both genders, with the mean apex of the kyphosis close to T6. The cervical lordosis tended to flatten with increasing age, particularly in males, with the cervical apex location shifting cranially. This association was significant in older males and females. The mean location of the cervicothoracic curve inflexion point moved from T3 towards C7–T1 with increasing age. Conclusion. The cervicothoracic spinal curvature undergoes progressive change through the lifespan with a subsequent cranial migration of the inflexion point between the thoracic kyphosis and cervical lordosis, accompanied by a similar shift in the cervical apex. Relevance Sensitive measures of spinal curvature have utility in determining changes attributed to age, deformity or trauma on cervicothoracic spinal alignment. The value of assessing the location of curve inflexion lies in the ability to quantify changes in the relationship between different regions of the human spine without problems associated with identifying specific vertebral landmarks. Ó 2002 Elsevier Science Ltd. All rights reserved. Keywords: Cervicothoracic junction; Thoracic; Cervical; Kyphosis; Spinal curvature; Sagittal spinal angles; Computer assisted measurement
1. Introduction Progressive change in adult spinal sagittal posture is a characteristic feature of the aging process, accentuated in women around the menopause through osteoporosis. Normal spinal alignment proposes that the mass of the
*
Corresponding author. E-mail address:
[email protected] (K.P. Singer).
head is balanced over the reciprocal primary and secondary curves of the spine, with a line of gravity bisecting the column through the transitional junctions (Fig. 1). The consequence of an accentuated thoracic curvature is mirrored in the cervical region with compensatory adjustments to head posture required to preserve forward gaze. These changes are represented in Fig. 2 which depict the alteration in spinal sagittal alignment over the life span. There have been few reports examining the magnitude of the changes to cervicothoracic
0268-0033/02/$ - see front matter Ó 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 2 6 8 - 0 0 3 3 ( 0 2 ) 0 0 0 3 0 - X
362
J.J.W. Boyle et al. / Clinical Biomechanics 17 (2002) 361–367
Fig. 1. Schematic representation of the normal spine which proposes the location of the line of gravity through the mass of the head, balanced over the sub-occipital, cervicothoracic, thoracolumbar, and lumbosacral transitional junctions.
spinal curvature and the implications of such change on postural mechanics of this region. Where radiography is employed to investigate the thoracic and cervical spine regions the clinical focus is the identification of active disease, fractures, recent trauma and related spinal pathology [1,2]. The cervicothoracic junction is however difficult to visualise adequately on routine clinical films [3] and this compounds the in vivo evaluation of alterations in spinal curvature in this region. Special ‘swimmer’s views’ of the cervicothoracic junction tend to induce an oblique orientation of the vertebral segments which further distorts the interpretation of change in spinal alignment. The measurement of the thoracic spinal curvature has been well documented [4–7]. Radiographic curvature analysis using a modification of the Cobb angle [6,8,9], pantographs [7] inclinometer [4], flexicurve [10] and computer assisted digitising [5,8,11,12] measures have defined, with some variability, the thoracic kyphosis. The variations in measurement results may, in part, be accounted for by the variations in measurement technique, particularly with Cobb angle calculation. The difficulty in visualising the upper thoracic vertebrae on chest films has lead to selection of other thoracic vertebrae as the upper measurement point. Measurement of the thoracic kyphosis from post-mortem radiographs
Fig. 2. Schematic depiction of the sequential adjustment in cervicothoracic spinal posture over the lifespan. An accentuated thoracic kyphosis tends to reduce the cervical lordosis leading to sub-occipital extension as a compensation to preserve forward gaze.
eliminates the difficulty in visualising the upper thoracic vertebrae due to the removal of surrounding bony and soft tissue structures. Using a retrospective survey Singer et al. demonstrated strong correlations for Cobb angle and computer assisted digitising measurement of the thoracic kyphosis from an in vivo and in vitro comparison of 22 cases [12]. This suggests that measurement of spinal curvature using post-mortem radiographs is a valid protocol for normative data reporting, however, no such study has been reported for the cervical region. Also Goh et al. reported that the computerassisted method was more appropriate when measuring thoracic kyphosis in cases with irregularities in their end-plate orientation [11]. Assessment of cervical spinal curvature has also been determined by radiographic [13–15], and surface contour techniques [16]. Refshauge et al. reported however that surface curvature could not be used to infer vertebral alignment due to the variability in the length of the spinous processes and overlying soft tissue [16]. As with the thoracic curvature different measurement techniques have been applied to assess the cervical curvature from
J.J.W. Boyle et al. / Clinical Biomechanics 17 (2002) 361–367
roentgenograms. Comparison of results thus becomes difficult. Cote et al. studied the reliability of measuring sagittal curves on lateral cervical spine radiographs [17]. They reported that angles obtained from C1 to C7 lacked clinical accuracy, however using the inferior endplates of C2 and C7 as the references for the cervical angle very high reliability was recorded. There is a very limited literature reporting the relationship between the cervical and thoracic curvatures. Refshauge et al. did describe a measure of the cervicothoracic angle, based on lines drawn through C4 and C7 and through C7 and T4 [16]. Consequently, the aim of this study was to: (1) quantify the cervical and thoracic curvatures of the human spine; (2) to determine the point of inflexion between the two regions and (3) to examine the influence of age on the cervicothoracic alignment.
2. Methods One hundred and seventy-two post-mortem lateral spinal radiographs were randomly selected from the archives of the Department of Neuropathology, Royal Perth Hospital. The spinal radiographs of the intact formalin fixed ligamentous spinal column were taken within 24 h of excision, and included 113 male aged 18– 90 years (mean age 48.3 years) and 59 females aged 18–92 years (mean age 54.6 years). Any case exhibiting post-mortem damage, advanced degenerative changes,
363
prior fractures/surgery, or deformity, were excluded from this survey. The anterior and posterior vertebral margins and the superior and inferior vertebral end-plates were traced for the second cervical to the twelfth thoracic (C2–T12) vertebrae directly from the lateral radiographs. For the modified Cobb angle assessment of the thoracic kyphosis, perpendiculars were extended from lines drawn through the superior endplate of T1 and the inferior end-plate of T12, with the angle measured at their intersection. A similar angle assessment was created for the cervical curvature by extending lines from the inferior end-plate of C2 and the inferior end-plate of C7. Additionally, a cervicothoracic junctional region angle measurement was taken using the superior vertebral end-plate of C6 and the inferior end-plate of T4, as represented in Fig. 2. The anterior and posterior vertebral margin tracings were then positioned on a digitising tablet (Digipad 5, GTCO, Rockville, MD, USA) which was linked to a PC. The two lines were digitised continuously to produce two arcs and the centre of each vertebral body denoted from C2 to T12. The arcs were smoothed to produce an averaged curve and a polynomial function was fitted to the resulting X–Y coordinates. The vertebral segment at the apex of the cervical and thoracic curvatures, and the vertebral level of inflexion between the thoracic and the cervical curvatures, was determined in software as the points of maximum and zero slope, respectively (Fig. 3).
Fig. 3. Representative ex vivo radiograph of the cervicothoracic spine (A) from which the Cobb angles; (B) for the cervical (C2–C7) and cervicothoracic junction (C6–T4), angles a & b, respectively, are defined. The digitising technique resolved an averaged curve from the vertebral body tracings from which cervicothoracic curvatures, points of maximum (apex ¼ A) and zero change (inflexion ¼ I), were derived in software from the polynomial curve fitting function; (C) for clarity, a hemisected spine was employed in this radiograph.
364
J.J.W. Boyle et al. / Clinical Biomechanics 17 (2002) 361–367
2.1. Statistical analysis Descriptive statistics were used to present cervical, cervicothoracic and thoracic Cobb angles and to record variations in the cervicothoracic curve inflexion and the cervical and thoracic apices. Additionally the three curvature variables were analysed for gender and age influences using two factor analysis of variance. As a structural change in curvature between regions was expected, Scheffe’s test was used for post hoc analysis between age groups and the Mann–Whitney U test for gender comparisons within age groups. The association between thoracic and cervical curvature in the older cohorts was assessed by linear regression. To test for intra-observer reliability a random selection of 18 unmarked films were measured on two separate occasions four weeks apart. For Cobb angle measures intra-class correlation coefficients (ICC) were computed. For the apical and inflexion segmental levels, the frequency of disagreement was recorded. A probability level of P < 0:05 was selected as the criterion for noting significant differences.
3. Results The intra-observer reliability of the Cobb angle measurement procedure resulted in ICCs of 0.98, 0.96 and 0.99 for the cervical, cervicothoracic, and thoracic angles, respectively. For the digitiser data, the frequency of disagreement for the cervical apex, cervicothoracic inflexion and the thoracic apex segmental levels was
