Research Review by Dr. Shawn Thistle©


June 2007

Study Title:

The association between cervical spine curvature and neck pain


Grob D, Frauenfelder H; Mannion AF

Publication Information:

European Spine Journal 2007; 16: 669-678.


Degenerative changes to the cervical spine, visible on radiographic examination, are known to be a normal part of the aging process. Observed changes can include facet joint space narrowing with sclerotic changes, uncinate process blunting or deformity, and disc space narrowing with or without osteophytic changes. These changes may or may not cause pain, and can be accompanied by changes in global or segmental lordosis of the cervical spine.

Many clinicians focus on these radiographic findings when explaining neck pain to patients, or when prescribing treatment regimens. Some go a step further, justifying and promoting extensive treatment plans or techniques to "prevent" progressive degeneration, deformity, and disability thought to occur from these changes. This exchange between doctor/therapist and patient occurs in the absence of quality evidence to support these claims.

Existing literature regarding degenerative spinal changes and clinical pain conditions has suggested the following:
  • degeneration shows a consistent (albeit weak) association with pain in the lumbar spine - a similar association has yet to be established in the cervical spine, as existing studies have yielded conflicting results
  • advanced imaging findings (MRI/CT) often do not correlate with clinical symptoms, prognosis, or risk of recurrence for spinal pain conditions (most of this literature is in the lumbar spine relating to disc pathologies)
  • degenerative changes in the cervical spine are often accompanied by reductions in anterior and posterior vertebral body height, which results in alteration of the cervical spine sagittal orientation (lordosis/kyphosis)
  • radicular pain can more often be explained by imaging findings such as osteophytes or foraminal encroachment when compared to general axial neck pain without radiation or radicular symptoms
  • there is general agreement that a lordotic cervical curvature represents the "normal" or “ideal” alignment, but literature regarding the normal value of the curve is varied (estimated to be between 20-35° when measured between C2 and C7), and is highly dependent on measurement method
The aim of this study was to evaluate whether changes in sagittal cervical spine orientation correlate with the presence and severity of axial neck pain in patients over 45 years of age. 107 volunteers (average age 68), were recruited from a patient population consulting an outpatient hospital clinic for lower extremity complaints (note: no current neck pain). Exclusion criteria were as follows:
  • previous cervical spine trauma
  • undergoing current treatment for neck pain
  • systematic disease involving the cervical spine
  • age under 45
Each patient was asked if they had experienced cervical spine pain within the previous 12 months. Those who replied yes were classified as the PAIN group and those who replied no were the NO PAIN group in the analysis. Lateral cervical radiographs were taken of each subject using an external guide to standardize the films as best as possible. External wires and guides were used to set up each patient to 20° inclination of the head (measured from horizontal to a line passing from the ear canal to the eye).

On the resulting films, the degree of cervical lordosis was measured using the posterior tangent technique - angles are measured between lines parallel with the posterior bodies of C2 and C7 for global, and between vertebral levels for segmental angles. As no standard exists for the normal value of cervical curvature, the authors incorporated established values for measurement error on x-rays to come up with the following system:

Total Curvature
  • Straight - -4°-+4°
  • Lordotic - less than -4°
  • Kyphotic - more than +4°
Segmental Curvature
  • Straight 0°-+4°
  • Lordotic - less than 0°
  • Kyphotic - more than +4°
At the time of investigation, a custom-made questionnaire was administered containing questions enveloping many aspects of chronic neck pain. The authors chose to create a questionnaire as no universally accepted, multidimensional tool exists in the German language (see below).

Pertinent Results:

  • the average age of the subjects in this study was 68 (47 male subjects, 60 female)
  • 54 of the 107 subjects indicated having neck pain in the previous 12 months, and hence formed the "pain group" - 53 indicated having no neck pain in the previous 12 months, and hence formed the "no pain group" - the only significant difference between the groups was the gender distribution (pain group: 73% female, no pain group: 40% female)
  • additional analyses were performed on males and females separately to account for the potential confounding effect of the gender difference in groups
  • the only significant difference between genders in the pain group was that women experienced more frequent headaches, and used non-operative treatment modalities more frequently
  • in the pain group - most had mild to moderate pain (average duration was 67 months but with a very wide standard deviation), with just under half experiencing radiating pain
  • there were no significant differences between the pain and no pain group for either total (C2-C7) or segmental curvature at any level
  • when genders were examined separately, no differences existed between pain and no pain groups for global curvature, or segmentally (except in men, C2/3 was slightly less lordotic and C6/7 more lordotic in the pain group)
  • in the pain group – there was no correlation between curvature and any clinical features of pain examined in the questionnaire
  • a segmental kyphotic deformity (>+4°) was found in at least one segment in 23% of the subjects in the pain group and 17% of the no pain group – the degree of kyphosis did not differ between groups
  • 2 radiographic anomalies were observed - one congenital fusion of C2/3 and one assimilation of the atlas - interestingly, both were in the no pain group

Conclusions & Practical Application:

Assessing sagittal curvature of the spine from x-rays is a controversial topic. There are entire treatment techniques based on this which rely heavily on pre- and post-treatment x-rays to evaluate progress. The validity of these techniques is questionable, and still needs to be elucidated. In addition, the most effective way to measure cervical curvature has yet to be determined, and measurement error is known to be high. It is also within reason to consider that many factors can affect cervical (or spinal) curvature while taking radiographs including pelvic tilt, sitting or standing, backrest shape when sitting, head positioning, etc.

In general, many manual therapists discuss "normal" radiographic degenerative changes, including changes in sagittal curvature, in the context of a patient's pain condition. The results of this study suggest that there is no association between sagittal alignment in the cervical spine and the presence of neck pain. This is in agreement with existing studies focusing on the lumbar spine.

There are some weaknesses of this study that deserve consideration when interpreting the results:
  • The subjective assessment was a questionnaire designed specifically for this study. The content of the form was adequately described in the paper and appears comprehensive, but it is curious why the authors did not simply rely on established outcome measures such as the VAS or NDI.
  • defining the pain group as those with neck pain in the last 12 months may not truly represent a neck pain group - future studies could include those that are seeking treatment actively for neck pain at the time of study
  • there were wide standard deviations in patient age and duration of pain, which limits the statistical power of this study for making conclusions about any specific patient population with neck pain
  • the authors point out: the fact that a group of individuals with or without neck pain do not show differences in the level of cervical lordosis does not mean that for an individual, a sudden change in curvature could not be important clinically
This study is a bit statistically weak, but suggests that sagittal cervical curvature does not correlate with neck pain in older patients. Although this should be considered when differentially diagnosing neck pain and communicating with patients, future well-designed studies are needed to clarify this issue. It is always important to consider each individual case, and as one of my favourite professors used to say: "Treat the patient, not the picture."