In the present study,
by utilizing the sagittal reconstruction images of CT chest,
abdomen and pelvis,
the prevalence of diffuse idiopathic skeletal hyperostosis was surprisingly higher than the previous plain radiograph based or CT based studies.
A previous CT based study conducted on Japanese population showed slightly lower prevalence as compared to our study [7].
However,
previous studies based on plain radiographs have reported much lower prevalence [3-6].
We believe that this high prevalence can be due to the fact that utilization of CT scans for evaluation of DISH resulted in proper and detailed evaluation of spine.
The gender distribution of DISH reported in this study is comparable to those reported by other authors as well i.e.
DISH occurs more frequently among men as compared to women [3-6].
Excellent agreement between the observers was seen in our study in classifying DISH according to criteria given by Resnick.
Previous CT based studies utilizing Resnick have reported a variable interobserver relationship ranging from moderate [7] to fair and good [16].
The agreement reported by Oudkerk et al [16] was moderate and excellent when a modified Resnick criteria,
designed solely for CT was applied.
We believe that a detailed evaluation of ligamentous ossification is possible by CT imaging and a good understanding of Resnick criteria could have led to this excellent agreement.
(Figure 2)
In this study,
majority of patients having DISH were not reported in the final report of radiology.
The radiological reporting rate of DISH has never been evaluated before.
This low rate of reporting DISH in the final report could be due to reluctance of radiologists to include this in the final report.
Reporting of DISH might have significant implications for the patient for e.g small vertebral fractures can become unstable and have significant associated neurological injury due to stiffness of the large ankylosed spinal column [17].
Therefore,
the importance of including the DISH in the reports needs to be stressed.
The results of our study showed that pelvic enthesopathy (Figure 3) is twice as common among patients with DISH as compared to non-DISH patients.
Previous study has also shown that pelvic enthesophytes generally are more frequently seen in DISH as compared to non-DISH [18].
Enthesophyte formation,
similar to osteophyte formation,
is considered as a response of skeleton to stress or repeated micro trauma [19] and is predominant in older patients,
specifically males [20].
This can be considered a reason of enthesopathy being common in our study because males comprised a significant portion of our study population with mean age around 60 years.
Formation of flowing osteophytes between two vertebral bodies is chronic process with approximate duration of ten years [21].
Therefore,
as suggested by Utsinger (22),
peripheral enthesopathies can be used to suggest the early diagnosis of DISH even when the involvement of spine is not present.
We found that Cobbs angle for thoracic kyphosis was significantly greater in patients suffering from DISH.
These cross-sectional results are comparable to those reported previously that patients with DISH have significantly greater thoracic kyphosis Cobbs angle [23,
24].
We hypothesize that spine morphology may be altered by ossification of anterior longitudinal ligament in DISH.
This may lead to reduced flexibility of anterior longitudinal ligament required to maintain an upright posture.
Thus the spinal curve is affected and there is limited ability to straighten the spine resulting in greater thoracic kyphosis in such patients.
However,
further studies are needed to investigate it.
During the course of this study we found that DISH was significantly associated with atherosclerosis (Figure 4).
We then incorporated the evaluation of atherosclerosis into our study.
We then found out that chances of DISH being common in patients with atherosclerosis are higher than in patients not having atherosclerosis.
This result contradicts with the previous report [25].
The previous study has reported that no consistent association exists between DISH and abdominal aorta calcification [26].
The reason for this contradiction could be that we utilized CT for evaluation of atherosclerotic calcification whereas in the previous study the abdominal aortic calcification was assessed by radiographs.
Due to improved and detail image evaluation by CT,
we believe that subtle calcifications can be detected and evaluated.
The present study demonstrates that lumbar spondylosis has a negative association with DISH.
This is opposite to what has been reported by Kagotani et al [4].
According to Kagotani et al lumbar spondylosis has significant association with DISH.
In our study,
patients with lumbar spondylosis were less likely to have DISH.
Therefore,
further studies are needed to determine the association of DISH with lumbar spondylosis.
Possibility of genetic factors may have a role in development of lumbar spondylosis and DISH.
This study reported that odds of having DISH are more than four times higher in the presence of diabetes mellitus and three times higher in the presence of hypertension.
Several studies have also assessed the relationship of metabolic factors with DISH.
Similar findings have been reported by other authors as well.
Kiss et al has also reported a strong association of DISH with diabetes [26].
Another study was conducted by Mader et al focusing on fasting blood glucose levels in DISH.
They found significant association between DISH and increased glucose levels [27].
Similarly,
another study has shown that individuals in their fifth decade may have DISH if they are obese,
have hypertension or diabetes in their first degree relative and have enthesopathies [28].
Few other studies have reported association of DISH with diabetes mellitus,
hypertension,
obesity and dyslipidemia [29,
30].
Therefore,
it can be assumed that although DISH is a musculoskeletal disease,
its link with factors of metabolic syndrome has made it important to be viewed as a systemic disease.
The results of this study can be viewed in light of several limitations.
Association of DISH with dyslipidemia could not be evaluated as patients did not undergo lipid profile examination for this study.
Similarly its association with obesity could not be evaluated.
We could not evaluate other sites of osteoarthritis such as knees,
hands,
etc as knee osteoarthritis has also been reported to be associated with DISH [4].
The patients were not followed to assess the longitudinal associations of the study variables.