Answer to the letter to the editor of Dr. D.E. Harrison et al. regarding 'The association between cervical spine curvature and neck pain' (D. Grob et al., Eur Spine J, published online Nov 18, 2006)

Autor: Dieter Grob, Anne F. Mannion, H. Frauenfelder
Jazyk: angličtina
Rok vydání: 2007
Předmět:
Popis: We thank Dr. Harrison for his interest in our work and for taking time to share his thoughts concerning the discrepancies between the findings of our respective studies. We apologize for wrongly characterizing one of his papers/inadvertently mixing up two of the papers [5, 6] in our Table 4 [3]. This was indeed an error on our behalf, and occurred as a result of the later insertion into the manuscript (which, incidentally, started out as an MD thesis written in German) of the wrong Harrison et al. reference; we hope that Harrison et al. will be prepared to interpret this as a compliment, in relation to the huge amount of work they have published in this field! To say that we suggested their results were “questionable due to improper study design” is somewhat exaggerated paraphrasing of the statements made in our discussion. We were merely looking for differences between the studies that might have accounted for the discrepant results. The notion that there may have been some methodological issues at stake was prompted by the observation that the mean cervical lordosis in their asymptomatic population was larger than the values previously reported in the literature for asymptomatics. We simply stated that Harrison et al. did not report whether identical radiographic methods (and in particular standardized subject positioning) were used in the two retrospective studies carried out 10 years apart; we do not say that comparable methodology was not used, only that it wasn’t mentioned in their papers (and this was later conceded by the authors themselves in their letter to the editor:”although our x-ray procedure was not detailed in our more recent paper…”), and was hence to be considered a possible explanation for the findings. Given the long time that elapsed between their two retrospective studies and the known advances in modern imaging techniques, we considered that this was a potentially plausible explanation, in the absence of any information suggesting the contrary. This was intended to constitute a normal process of scientific deduction, not a pointed criticism of their work. We already discussed in our original paper [3] the potential difference between the two study groups arising from Harrison et al.’s exclusion of individuals with segmental/total kyphosis [7], and we confirmed that applying the same criterion to our own subject group did not alter our findings. The other exclusion criteria used by Harrison et al. [7] but not by ourselves [3], i.e., moderate–severe disc disease, and any significant forward head posture, may indeed have contributed to their higher values for the cervical lordosis in their NO PAIN group. However, one would then have to argue that, using these exclusion criteria, one no longer has a valid representation of the asymptomatic population. In other words, it may explain the differences between the cervical lordosis in two select groups of people, but it doesn’t answer the question as to whether lordosis and symptoms per se are related. This was the question we sought to answer. The fact that we were examining older individuals (mean 68 years of age, some of whom would naturally have had degenerative changes) may or may not have accounted for the differences between our studies for the curvature of the NO PAIN group. Whilst it is an explanation that is worthwhile considering, it is by no means as unequivocal as Harrison suggests in his statement “the exclusion of moderate to severe degenerative joint disease would naturally increase the mean cervical lordosis”. A number of studies have shown that the cervical lordosis increases with age [2, 4, 10]; however, whether this increase is less marked in the presence of degenerative changes [2] or not [4] remains unclear. In the study cited [2] in Harrison’s letter, the 10° difference in the mean lordosis in patients with disc space narrowing (one measure of degeneration) was seen only in patients over 50 years of age; in the younger age-groups there was no relationship between cervical lordosis and the severity of degenerative changes (assessed as disc space narrowing, end-plate sclerosis, or anterior and posterior osteophyte formation). In our study, degenerative changes were not assessed, and so we cannot pass any comment on the relationship between the degree of degeneration and cervical lordosis in our patients; neither can Harrison, and hence his belief that “the controversial findings of Grob et al. compared to ours and other reports in the literature is mostly accounted for due to their specific senior adult population having moderate–severe degenerative disc disease” must also be acknowledged as mere supposition. Importantly, the main reason for comparing our data with Harrison et al.’s was actually in relation to the lack of difference between the PAIN and NO PAIN groups, not just because of the difference between our individuals with NO PAIN and their individuals with NO PAIN. In this sense, since it was as likely (or as unlikely) that there were as many individuals with degenerative changes in both our PAIN and NO PAIN groups, the presence of degeneration in itself is not an explanation for the lack of difference between the respective studies. We have already conceded that the difference in patient type (non-consulting, and with non-specific neck pain) might be partially responsible, although we considered this unlikely since none of the pain characteristics (severity, frequency, etc.) that we measured showed any relationship whatsoever with cervical spine curvature. Incidentally, Harrison’s statement that 41–46% of our subjects had signs of cervical radiculopathy is inaccurate: radiating pain does not necessarily mean neuropathic radiation (radiculopathy); it is commonly myofascial radiation. We thank Harrison et al. for highlighting further literature on the topic. The McAviney [8] article was indeed relevant to our study, but was unfortunately published after our study had been written up. We failed to systematically search the literature again before finally submitting the paper, and hence did not locate this one. The Nagasawa et al. paper [9] was not considered to be of direct relevance to our research on neck pain; it was out of the scope of this study to start delving into the tension–headache literature. Finally, with regards to the history of the posterior tangent method of measurement, Albers [1] was actually the first to describe this technique, as we correctly mentioned in our original paper [3]. If we were inaccurate in citing who came after him, and in what order, we apologise for this and thank Harrison et al. for providing the interested reader with the detailed historical account. If the volunteers declared having NO PAIN, they were not requested to complete the PAIN questionnaire; hence, we are unfortunately unable to provide any further information regarding complaints other than neck pain. In summary, we don’t think there is any dispute that we examined a different age-group (older individuals) and patient type (people who were not specifically consulting for neck pain) compared with Harrison’s studies: this was indeed clear from the study description and from our discussion. However, this does not mean that these differences are necessarily the explanation for the discrepant findings between our study and that of Harrison et al. [7]. Further, it does not invalidate our own conclusion that “abnormalities” of the sagittal profile observed in the older patient with neck pain must be considered coincidental, i.e. not necessarily indicative of the cause of pain, and that this should be given due consideration in the differential diagnosis of patients with non-specific neck pain”.
Databáze: OpenAIRE
Externí odkaz: