Correlation between kinematic sagittal parameters of the cervical lordosis or head posture and disc degeneration in patients with posterior neck pain: retrospective cross-sectional study

The aim of this study was to examine degrees of the cervical disc degeneration and the parameters of cervical sagittal balance in plain radiographs, representing cervical lordosis or head posture in subjects with posterior neck pain. This study was a retrospective cross-sectional study that analyzed 113 patients with posterior neck pain with or without radiating pain. The kinematic sagittal parameters of cervical radiographs were obtained at the occipito-cervical (O-C2) angle, sagittal Cobb’s angles of C1-C2, C2-C7, and sagittal vertical axis (SVA) of C1-C7 and C2-C7. Cervical disc degeneration was evaluated using the sum of Prrmann grades and total modied Matsumoto scores calculated from magnetic resonance images (MRI) of the cervical spine. Pearson’s correlation analysis was conducted to evaluate the relationship of these parameters and a linear regression method was used to determine associations between sagittal cervical alignment and disc degeneration.


Abstract Background
The aim of this study was to examine degrees of the cervical disc degeneration and the parameters of cervical sagittal balance in plain radiographs, representing cervical lordosis or head posture in subjects with posterior neck pain.

Methods
This study was a retrospective cross-sectional study that analyzed 113 patients with posterior neck pain with or without radiating pain. The kinematic sagittal parameters of cervical radiographs were obtained at the occipito-cervical (O-C2) angle, sagittal Cobb's angles of C1-C2, C2-C7, and sagittal vertical axis (SVA) of C1-C7 and C2-C7. Cervical disc degeneration was evaluated using the sum of P rrmann grades and total modi ed Matsumoto scores calculated from magnetic resonance images (MRI) of the cervical spine. Pearson's correlation analysis was conducted to evaluate the relationship of these parameters and a linear regression method was used to determine associations between sagittal cervical alignment and disc degeneration.

Results
A signi cant positive correlation was found for the C2-C7 angle using the sum of the P rrmann grades and total modi ed Matsumoto scores. The O-C2 angle was negatively correlated with the sum of the P rrmann grades and total modi ed Matsumoto scores. The C1-C2 angle was also negatively correlated with the sum of the P rrmann grades. No signi cant relationship was found between the cervical SVA and cervical disc degeneration.

Conclusions
The sagittal cervical parameters representing cervical kyphosis and jaw lifting posture, were found to be signi cantly correlated with the degree of cervical disc degeneration. These ndings suggest that the loss of the natural sagittal lordosis of the cervical spine may contribute to the progression of disc degeneration, rather than the forward head posture.

Background
Loss of cervical lordosis is the most common disorder of sagittal cervical balance. [30,33] Although the sagittal alignment of the cervical vertebrae can vary with age and sex, [8,30] [19]. The estimated prevalence of loss of cervical lordosis in patients with posterior neck pain is approximately 42%. [13] Recently, several studies have demonstrated that a spectrum of cervical disorders are associated with a loss of cervical lordosis, i.e. kyphosis. [10,16,30,32] A cross-sectional study revealed that decreasing natural cervical lordosis was correlated with increasing Neck Disability Index scores in pre-operative subjects. [14] Another prospective cohort study reported that patients with higher preoperative lordotic angles showed better outcomes than those with a kyphotic alignment. [32] Some studies have shown negative consequences of cervical malalignment on the health-related quality of life. [1,2,30] Forward head posture (FHP) is common in sagittal cervical imbalance in patients with symptomatic neck pain. FHP is usually de ned as an increased sagittal vertical axis (SVA) in which the head is shifted anteriorly to the shoulder plane compared to the neutral posture. The normal physiological C2-C7 SVA is estimated to be 16.8 ± 11.2 mm in asymptomatic subjects. [11] Several studies have found that an increased SVA value is related to clinical symptoms. [3,34] The majority of research on sagittal cervical balance has been focused on analyzing radiographs of the cervical spine and comparing them with clinical symptoms, [3,17,30] whereas few studies have focused on the relationship between cervical sagittal alignment and disc degeneration as seen on magnetic resonance imaging (MRI). Previously, the progression of cervical disc degeneration was higher in the nonlordotic group than in the lordotic group; however, the study only analyzed the C2-C7 Cobb's angle regardless consideration of the other sagittal variables, such as SVA or the occipito-cervical angle. [26] The results of this study may serve as investigation of the relationship between cervical disc degeneration as assessed using MRI and parameters of sagittal cervical balance in plain radiographs, representing cervical lordosis or FHP in patients with posterior neck pain.

Subjects
Four hundred and twenty-one patients with posterior neck pain who visited the spine center of a university hospital in a metropolitan area were retrospectively analyzed. Only subjects with both cervical spine radiographs and cervical MRIs were included. The exclusion criteria were as follows: history of cervical spinal surgery or trauma, age under nineteen years, symptoms or signs of in ammatory back pain, such as ankylosing spondylitis, imaging evidence of concurrent myelopathy, and spine fracture. A total of 113 patients were included in our study population.

Cervical disc degeneration measured by MRIs
3.0-T cervical MRIs (Discovery MR750; GE Healthcare, Milwaukee, WI) were performed to evaluate disc degeneration. Sagittal and axial T2-weighted MR imaging was performed for each cervical level. The severity of disc degeneration was assessed according to the P rrmann grades [28] and modi ed Matsumoto classi cation scores [25]. Each grade and score of the cervical levels from C2-C3 to C6-C7 was measured, and summation analysis was performed. P rrmann grades are used to evaluate degenerated intervertebral discs for (i) distinction of the annulus and the nucleus, (ii) disc structure, (iii) signal intensity of discs, and (iv) height of discs by using T2-weighted mid-sagittal images. [28] The modi ed Matsumoto classi cation scores assess disc degeneration for (i) change in the signal of the disc, (ii) posterior protrusion of the disc, and (iii) narrowing of disc space in T2-weighted axial and midsagittal images. [25]

Lordotic and Non-lordotic groups
To identify the speci c correlation of kinematic sagittal parameters with cervical disc degeneration, the patients were divided into two groups (lordotic group and non-lordotic group), according to the C2-C7 Cobb angle observed on cervical plain radiographs. We set the cutoff of the C2-C7 lordotic angle as -17.17°, based on the study by Harrison et al. [12] 3. Statistical analysis The comparisons between groups were performed using an independent t-test. The reliability of the measurements was assessed by examining the intra-observer and inter-observer reliability with intra-class correlation coe cients. The correlations between cervical parameters were determined using a Pearson's test. A linear regression method was used to determine associations between the parameters of sagittal cervical alignment and the degree of cervical disc degeneration. All statistical analyses were performed with SAS version 9.3 (SAS Institute, Cary, NC).

Ethics statement
The present study protocol was reviewed and approved by the Institutional Review Board of Gangnam Severance Hospital, Yonsei University Medical College (# 2016-0558-001).

Basic characteristics of the two groups
This study included a total of 113 patients (55 males, 58 female). The baseline characteristics of the participants are summarized in Table 1. Twenty-six patients (10 males, 16 female) were classi ed into the lordotic group, and 87 patients (45 males, 42 female) were classi ed into the non-lordotic group. There were no statistical differences in age and sex between the two groups.

Reliability Analysis
The intra-rater and inter-rater observer agreements for cervical parameters and disc degeneration were analyzed ( Table 2). In both intra-rater and inter-rater studies, good agreements were found, with κ values ranging from 0.72 to 0.91.

Analysis of cervical angles
The C2-C7 angle was signi cantly correlated with the O-C2 and C1-C2 angles. The O-C2 angle was signi cantly correlated with the C1-C2 angle, C1-C7 SVA and C2-C7 SVA. The C1-C7 SVA and C2-C7 SVA were correlated with the C1-C2 angle (Table 3). There was no signi cant correlation between age and any measured radiographic parameter.

Analysis of cervical disc degeneration by P rrmann grades
The degree of disc degeneration was assessed using P rrmann grades at each level. In the assessment of disc degeneration according to P rrmann grades, grade I was most commonly observed at the C2-C3 level. Grade V was more common at the C5-C6 level than at any other level (Fig. 2). 5. Correlation between kinematic parameters of the cervical spine and disc degeneration 5.1. Cervical angles and disc degeneration Using a multivariate regression analysis adjusted for age, the C2-C7 angle revealed a signi cant positive correlation with both the sum of P rrmann grades (r = 0.33, P < 0.005) and the total modi ed Matsumoto scores (r = 0.27, P < 0.005), although age was not a co-linear factor (Fig. 3a, 3b). There was a signi cant negative correlation between the O-C2 angle, and both the sum of P rrmann grades (r = -0.16, P = 0.005) and the total modi ed Matsumoto scores (r = -0.11, P = 0.02) (Fig. 3c, d). The C1-C2 angle was negatively correlated with the P rrmann grades (r = -0.15, P = 0.018) (Fig. 3e).

SVA and disc degeneration
The C2-C7 SVA and the C1-C7 SVA did not reveal a correlation with the sum of P rrmann grades or the total modi ed Matsumoto scores (Table 4). In both subgroups (lordotic and non-lordotic groups), the SVA had no signi cant correlation with the P rrmann grades or the total modi ed Matsumoto scores (Table  4).

Discussion
The cervical spine naturally maintains a lordotic curvature to compensate for the thoracic kyphotic curvature. [8] As the loss of cervical lordosis progresses, the deformity also tends to progress rapidly by producing abnormal forces to the head and neck. [2,30] Even with mild sagittal imbalance, detrimental symptoms can develop, which worsen with sagittal imbalance progression. [9] The vertebral disc is designed to maintain an isotropic form by transmitting axial load uniformly across the disc and vertebral endplate. [22,23] In cervical spinal positions, such as extension, exion, or lateral bending, the load of the disc is transmitted uniformly over the endplates. [23] Loss of cervical lordosis may alter this isotropic nature of disc loading and consequently contribute to continuous irregular loading, which accelerates disc degeneration. [23,31] This degeneration can be aggravated by normal aging, calci cation of the endplate, or decreased peripheral blood supply. [5] Abnormally increased mechanical pressure has also been shown to reduce the nutritional support of the disc and lead to disc degeneration. [29] Our results demonstrate a correlation between loss of cervical lordosis and cervical disc degeneration. In this study, cervical disc degeneration was evaluated using both P rrmann grades and the modi ed Matsumoto scoring system, whereas most previous studies have only used either of them. [4,15,25,26] P rrmann grading assesses the homogeneity of disc structure and includes a distinction between the annulus and nucleus, whereas the modi ed Matsumoto scoring system considers the degree of the posterior disc protrusion and narrowing of disc space without considering the homogeneity of the disc. This difference in the evaluation criteria may explain the different relationship between the C1-C2 angle and disc degeneration, which the C1-C2 angle is only correlated with the P rrmann grades.
Changes in sagittal cervical alignment, such as FHP, may cause or result in adaptive mechanisms to global alignment change, which affects all spinal levels (including the cervical, thoracic, and lumbar regions). [18,30] Contractions of the neck muscles due to vestibulocollic or cervicocolic re exes induce anterior shifting of the head/neck center of gravity, resulting in a change in the spinal alignment. [27,36] These re exes cause cervical muscle spasm, representing shortening of the posterior neck extensor muscles and the tightening of the anterior neck muscles, which may increase the SVA. [6] Previous studies have shown that a larger C2-C7 SVA is related to higher Neck Disability Index scores [35] and demonstrated a correlation between the C2-C7 SVA and the C1-C2 angle. [35] The results of this study also showed that the C2-C7 SVA was signi cantly correlated with the O-C2 angle and the C1-C2 angle. However, to date, the clinical consequences of increased cervical SVA on cervical disc degeneration have not been described. In this study, the change in SVA was not correlated with cervical disc degeneration. Both in the lordotic and non-lordotic groups, the SVA was not correlated with cervical disc degeneration.
These ndings indicate that the SVA, which has been suggested to be related to FHP in previous studies, has little effect on disc degeneration. Cervical disc degeneration progressing with normal aging could be aggravated by the loss of natural sagittal angles rather than increased cervical SVA. Recently, one study showed signi cant correlation between cervical lordosis and C2-C7 SVA in asymptomatic Chinese population.
This may mean that the rate of disc degeneration is greater than that of FHP when clinical symptoms occur.
The optimal position and angle of the occipital bone and the cervical axis have been topics of discussion. [20,21,24] In this study, there was a negative correlation between the O-C2 and C1-C2 angle and the C2-C7 angle, as well as a signi cant correlation with the SVA. These results suggest a correlation between the occipital-cervical axis, cervical kyphosis, and FHP. There was a signi cant correlation between the occipito-cervical angle and cervical disc degeneration in this study. A recent cohort study showed that an increased occipito-cervical angle may result in large biomechanical stress on the adjacent structures or deformation of cervical alignment. [21] A previous study has shown that the loss of the natural C2-C7 angle facilitates cervical disc degeneration. [26] In addition, our ndings suggest that a more negative occipito-cervical angle may accelerate disc degeneration. In summary, the SVA, the occipito-cervical angle, and the loss of cervical lordosis, expressed easily as the "forward head posture", "the jaw lifting posture" and "the cervical kyphosis", were correlated with each other. Also, only the latter two, the jaw lifting posture and the cervical kyphosis had were correlated with disc degeneration. Although the causeand-effect relationship is unknown, it can be interpreted that FHP worsens the jaw lifting posture and cervical kyphosis, which may cause disc degeneration.
This study has some limitations. The pathophysiological mechanism of disc degeneration due to the loss of cervical lordosis remains unknown. As the analyses were cross-sectional, determining a cause-andeffect relationship is di cult. Furthermore, neck pain was not classi ed as causal factor in this study. As cervical disc degeneration progresses, neck pain or uneven loading to the cervical disc can induce deformity of the sagittal alignments, including both cervical and occipito-cervical angles. Long-term changes in the sagittal cervical parameters and disc degeneration were also not evaluated. Prospective longitudinal studies with long-term follow-up are necessary to investigate the clinical implications and the interactions between the alignment of cervical spines and the discs. Finally, as we did not include an asymptomatic group of participants, the results may not be generalizable to whole populations. Future long-term longitudinal studies in a general asymptomatic population are needed.

Conclusions
This study showed that sagittal cervical parameters, such as the C2-C7 angle, O-C2 angle, and C1-C2 angle, were correlated with the degree of cervical disc degeneration in patients with posterior neck pain. However, there was no signi cant correlation between the SVA and disc degeneration. These results revealed that loss of the natural cervical lordosis is correlated substantially with cervical disc degeneration, rather than the FHP. and edited the writing. HJ was a major contributor in writing this manuscript and JH conceptualized the topic and supervised the study. All authors read and approved the nal manuscript.

Funding
Not applicable.

Availability of data and materials
We declare that the materials described in the manuscript will be freely available to all scientists for noncommercial purposes.
Ethics approval and consent to participate Ethics approval was provided by Gangnam Severance Hospital, Yonsei University Medical College and the consent of patients was not applicable in this study.

Consent for publication
Not applicable.

Competing interests
All the authors in this manuscript have read and approved the nal version submitted, and no author has a con ict of interest in this submission.    Correlation between sagittal angles of the cervical spine and the degree of cervical disc degeneration.