Diffuse Idiopathic Skeletal Hyperostosis
(DISH) in the dog.
A retrospective radiographic cohort study.
Onderzoeksstage verslag door: Drs. H.C. Kranenburg
Begeleiders: Dr. B.P. Meij
Prof. Dr. H.A.W. Hazewinkel
Drs. L.A. Westerveld
I DISH in humans 3
1.1 History and terminology 3
1.2 Diagnostic criteria 3
1.3 Differential diagnosis 4
1.4 Prevalence 4
1.5 Distribution in the skeleton 5
1.6 Etiology 6
1.7 Clinical symptoms 7
1.8 Treatment 8
II DISH in dogs 8
2.1 Canine anatomy 8
2.2 DISH in dogs 9
2.3 Spinal skeleton of a dog with lumbar DISH 10
2.4 CT and MRI of a dog with DISH 11
III DISH versus spondylosis deformans 12
3.1 Spondylosis deformans in humans 12
3.2 Spondylosis deformans in dogs 12
IV Aim of the study 13
Materials and Methods 14
Summary and Conclusion 24
Study design: A retrospective radiographic cohort study.
Objective: To determine the prevalence of diffuse idiopathic skeletal hyperostosis (DISH) in an out-patient
population of dogs (age > 1 year).
Summary of Background data: DISH is a systemic disorder of the axial and peripheral skeleton which is
common in humans but still unrecognised in dogs. The etiology of DISH in humans is unknown and the
condition occurs in middle aged and elderly human patients. It results in ossification of soft tissues such as
longitudinal spinal ligaments, joints, sites of attachment of tendons, muscles and/or capsules to bone. Symptoms
can vary from mild back pain and little stiffness to more serious symptoms such as neurologic deficits and spinal
fractures. The disorder is incidentally described in veterinary literature in dogs and may be difficult to
distinguish from spondylosis deformans.
Methods: Radiographs of 2041 dogs over 1 year of age, visiting the University Clinic for Companion Animals
in Utrecht between February 2003 and January 2008, were reviewed for DISH in the spine, using the criteria
stated by Resnick et al (1976). Four or more – level involvement was defined as DISH. Three-level involvement
was defined as pre-stage DISH and recorded separately.
Results: The overall prevalence of DISH in a cohort of 2041 dogs was 3.8%. The prevalence of DISH increased
with age; in dogs over 7 years the prevalence of DISH was 5.7%. In male dogs this was 7.1% and in female dogs
4.4%, resulting in a male dog-female dog ratio of 1.6 to 1. Boxers were most frequently affected by DISH; the
prevalence in Boxers over 7 years old was more than 50%.
Conclusions: DISH seems to be a prevalent disorder in dogs. The overall prevalence in this animal out-patient
population is 3.8%. Boxers are most commonly affected, which may indicate a genetic susceptibility for DISH.
Key words: diffuse idiopathic skeletal hyperostosis, DISH, dogs, prevalence, Boxer, spondylosis deformans.
I. DISH in humans
Diffuse or disseminated idiopathic skeletal hyperostosis (DISH) is a common, systemic disorder of the axial and
peripheral skeleton in humans. The etiology of DISH in humans is unknown and the condition occurs in middle
aged and elderly patients. It results in ossification of soft tissues such as longitudinal spinal ligaments, joints,
sites of attachment of tendons or muscles and capsules to bone.
1.1 History and terminology
DISH was first comprehensively described by Forestier and Rotes-Querol in 1950. Although other authors had
previously mentioned anatomic changes almost similar to the descriptions by Forestier and Rotes-Querol, the
publication of the latter was the first clinical and radiological study of this disorder. Forestier and Rotes-Querol
differentiated DISH (which they named ankylosing hyperostosis) from ankylosing spondylitis and spinal osteo-
DISH is not a typical disorder of recent history. Signs of DISH were found for example in the skeletons of
dinosaurs, Saber toothed cat, in the remains of Rhamses II (1302-1213 BC in Egypt) and in ancient clergymen in
the Netherlands.2, 3, 4, 5
Over the years the disorder has been given many other names, including monoliform hyperostosis, spondylitis
ossificans ligamentosa, hyperostotic spondylosis, hyperostosis of the spine, ankylosing hyperostosis, Forestier‘s
disease, generalized juxta-articular ossification of vertebral ligament and spondylosis hyperostotica. 6 The current
name used in medical literature is diffuse idiopathic skeletal hyperostosis or DISH which was introduced by
Resnick et al in 1975, who was the first to acknowledge the extraspinal manifestations of DISH.7
1.2 Diagnostic criteria
Resnick and Niwayama (1976) also postulated three diagnostic criteria for the diagnosis of DISH in humans:
1. The presence of flowing calcification and ossification along the anterolateral aspects of at least four
contiguous vertebral bodies with or without localized pointed excrescences at intervening vertebral
2. The relative preservation of disc height in the involved areas and the absence of extensive radiographic
changes of degenerative disc disease (intervertebral osteochondrosis), including vacuum phenomena and
vertebral body marginal sclerosis.
3. The absence of apophyseal joint bony ankylosis and sacroiliac joint erosion, sclerosis, or intra-articular
For dogs, the terms anterolateral and disc height are described as venterolateral and intervertebral disc space,
Utsinger (1985) suggested revising the Resnick criteria for epidemiological purposes. With these criteria it
would be possible to include early stages of DISH. He postulated three criteria:
1. Contiguous ossification along the anterolateral aspect of at least four contiguous vertebral bodies,
primarily in the thoraco-lumbar spine. Ossification begins as a fine ribbon-like wave of bone but
commonly develops into a broad, bumpy, buttress-like band of bone.
2. Contiguous ossification along the anterolateral aspect of at least two contiguous vertebral bodies.
3. Symmetrical and peripheral enthesopathy involving the posterior heel, superior patella or olecranon,
with the entheseal new bone having a well-defined cortical margin.
Three categories of DISH could be made according to these criteria:
A. Definite DISH: criterion 1
B. Probable DISH: criterion 2, 3
C. Possible DISH: (i) 2 and 3
(iii) 3 (particularly if calcaneal spurs occur together with olecranon or patella
He added two criteria for excluding patients with advanced spondylosis deformans:
1. Abnormal disc space height in the involved areas.
2. Apophyseal (AP) joint ankylosis.6
The difference between category B and C (i) is not clear. Furthermore, the two exclusion criteria is not sufficient
when spondylosis deformans is interpreted as: ‗a degenerative disease of the spine characterised by osteophytes
forming a spur on the vertebral bodies originating from the end-plate region‘.9 A distinction on the origin of the
spur or bridge should be made, but is not easy on radiographs.
Fornasier et al (1983) investigated 296 thoracic spines for changes which could be defined as early DISH.
According to the histopathological abnormalities three classes were described. Early changes consisted of
flattened spurs of bone originating from the waist of the vertebral body at the entheseal insertion into the cortical
(i) this extends adjacent to, but distinctly separate from, the vertebral body (except at the entheseal insertion) in
two separate body levels, it was called Type I;
(ii) this plate extends across the intervertebral space level separate from any disc or annulus fibrosus calcification
and from osteophytes arising from the vertebral body, it was classified Type II; and
(iii) when ossification extended overlapping the intervertebral disc space resulting in confluence of the new bone
from the adjacent vertebral body, it was named Type III. 10
1.3 Differential diagnosis
Based on the Resnick criteria DISH can be differentiated from a few other spinal disorders. 8 In typical
spondylosis deformans a flowing pattern along at least four contiguous vertebral bodies is not found and
ossification in the spine originates from the vertebral body instead of the longitudinal ligament.8 The
preservation of the disc space is not present in intervertebral osteochondrosis and the changes across apophyseal
and sacroiliac joints are characteristically seen in ankylosing spondylitis. 11 In DISH, the syndesmophytes project
ventrally from the vertebral bodies with the classic appearance of flowing candle wax, forming an extra-articular
ankylosis. They can be distinguished from the more cranio-caudally oriented ‗bamboo spine‘ outgrowths that
form an intra-articular ankylosis in ankylosing spondylitis. Ankylosing spondylitis usually starts in late
adolescence and early adulthood and consists of inflammatory spinal pain and stiffness, decreased range of
motion and after many years it can result in characteristic postural abnormalities such as ‗the Bechterew stoop‘
(= marked thoracic kyphosis). The presences of degenerative signs like facet hypertrophy and disc space
narrowing usually exclude the diagnosis of DISH.12 Some authors describe DISH as a variant of osteoarthritis
(OA), but without the degenerative signs of sclerosis, loss of intervertebral disc space and degenerative facet
joint changes, as seen in classic OA.13 DISH may be a distinct clinical entity.15, 16, 17 Others in the 1960‘s have
classified it as a special form of spondylosis deformans.18
DISH is mostly seen in the elderly and demonstrates a male predominance. 11, 12, 14 The incidence increases with
body weight in both genders.6, 11 The prevalence in man can vary around the world.11 In a hospital population in
the USA of people over 50 years old, a prevalence of 25% in males and 15% in females was found. 19 DISH was
found to be less common in African blacks, Afro-Americans, Native Americans, and Asians than in Caucasians
living in the USA.13, 19, 20 However, in Pima Indians living in the Gila River reservation in Arizona, USA, a very
high incidence of DISH and diabetes mellitus was found in 1973 by Henrard and Bennett. 6 In Korea, a
prevalence of 5.4% in males and 0.8% in females was found using the Resnick criteria. When looking at two or
more bridges, the so-called Julkunen‘s criteria, the prevalence increased to 7.1% in males and 3.2% in females. 21
In Israel, a prevalence of 9.8% was found in a cohort of 1020 humans over 45 years of age. 22 Overall, the
different prevalences found throughout the world suggest a possible genetic or ethnic factor in the prevalence of
In The Netherlands, a recent study of the human patient population from a clinic for internal medicine, showed a
male prevalence of 22.7% and 12.1% in females. In this study the authors also looked at ossification of the
anterior longitudinal ligament over three, in stead of four, contiguous vertebral bodies. This was considered to be
a precursor of full blown DISH. It was defined as ‗pre-stage DISH‘ and recorded separately. This pre-stage
DISH was found in 4.6% of the patients and more frequently in females. 23 Weinfeld et al (1997) made that same
differentiation but called it ‗likely dish‘ and ‗strictly dish‘. 19 In a study of 635 persons in Hungary, a prevalence
of 6.1% in males and 1.2% in females was found. When using the modified Resnick criteria; looking at two or
more bridges, prevalences of 27.3% in males and 12.8% in females were found. 14 In a cohort of Italian females a
prevalence of 15.1 % was reported.24
1.5 Distribution in the skeleton
The portion of the spine that is typically involved in humans is the thoracic region (Figure 1a and 1b). 11 Even in
patients with cervical or lumbar complaints the radiographic abnormalities are often found on the thoracic
spine.6, 12, 16 The syndesmophytes in human patients with DISH are usually found at the right side of the spine in
the thorax, presumably because of the protecting pulsatile effect from the aorta on the left side. Patients with
DISH and situs inversus show more ossification at the left side of the thoracic spine. 12 In humans it was found
that the thoracic abnormalities were most frequent in the 7 th to 10th thoracic vertebrae.6, 8 Mata et al (1993)
showed that the technique of human chest radiographs yields a sensitivity of 77%, specificity of 97%, positive
predictive value of 91% and a negative predictive value of 91% for diagnosing DISH in humans. 25
Fig. 1. Thoracic anterior-posterior radiograph (A) and CT reconstruction (B) of a human with DISH.
In humans, DISH in the cervical region is less common but not rare. 6, 26, 27 Mechanical dysphagia, sleep apneu
and difficulty with intubation are complications associated with cervical DISH. If the syndesmophytes are
impinging on anterior structures, surgical resection can be an option. 12
In the lumbar region radiographic changes resemble those in the thoracic spine but without the predilection of
the right side (Figure 2). Human patients with lumbar DISH often have changes in the cranial part of the lumbar
region. The vertebral ossification of the ligament may be as much as 2 cm thick. 6, 12
Fig. 2. Lumbar DISH in a human patient (lateral view).
Extraspinal manifestations of DISH are no exception.6, 18 Some authors find this so common that it should be
included into the diagnostic criteria.6 Various anatomic locations, such as joints, sites of attachment of ligaments,
tendons and capsules to bone can be affected. In humans, every location has its own characteristic findings.
These are usually bilateral and symmetrical.12
The etiologic factors are still not clear. Various metabolic, endocrinological, genetic and environmental factors
have been postulated but none has yet really been proven. 17, 28 DISH is often linked to obesity.1, 6, 8, 12, 15, 17, 22, 27,
29, 30, 31
In a case control study of 131 patients the body mass index (BMI) was significantly higher in the DISH
group than in the control group.6 Clinical consequences of obesity can be hypertension, osteoarthritis, pulmonary
and cardiac failure, diabetes mellitus and cardiovascular diseases. It is still not certain if the cardiovascular risk
factors associated with DISH are due to this disease or due to obesity. 22 Vezyroglou et al (1996) found that
differences in metabolic abnormalities persisted even after adjustment was made for BMI (body mass index).30
Some authors suggested that hyperglycemia was the most useful laboratory abnormality concurrent with DISH. 6
Others found no significant difference in glucose levels between patients with and without DISH. 17, 31 The
relationship with Diabets Mellitus type II (not-insulin dependent DM) and DISH was often postulated, but
remains controversial.15, 16, 28 Different authors disputed this relationship and found no differences in the
prevalence of DM between 50 patients with DISH and a control group of 50 persons without DISH. 32 In a study
of 133 patients with DM and a control group of 133 persons no statistically significant difference was found
between the two groups in the prevalence of DISH. 29 Vezyroglou et al (1996) found that DM alone was not a
risk factor, but in combination with high levels of uric acid and/or hyperlipemia, patients had a significantly
higher incidence of ankylosing hyperostosis.30
DISH has been related to abnormal bone cell growth or activity which could reflect the influence of metabolic
factors that lead to new bone formation. For example serum matrix Gla protein could be a marker of
osteometabolic syndromes, like DISH, which causes hyperostosis (Figure 3).16 Some authors suggested that
hypervascularity could be the localizing factor.16, 27 Others suggested that hyperinsulinemia, associated with high
BMI, suppressed the production of insulin-like growth factor (IGF) binding protein-1. As a result, it aggravated
the growth-promoting effect of IGF, which in turn may induce hyperostosis. 22 Denko et al (1994) reported that
DISH patients had elevated insulin and growth hormone values.31 It was suggested that hyperinsulinemia may be
the factor that link metabolic parameters with the development of hyperostosis. 6, 15 It was also postulated that
heavy work may correlate with the extent of DISH involvement. 33
In summary, the pathogenesis is not clear; many explanations for the onset and development of hyperostosis of
ligaments are hypothetical and lack strong and significant evidence for support. Further investigation is
Fig. 3. Possible pathogenetic mechanism driving bone deposition. 16
1.7 Clinical symptoms
DISH is known to affect the middle aged and elderly and is often asymptomatic or only associated with mild
back pain, little stiffness and some restriction in spinal range of motion. 12, 14, 34 Usually symptoms are relatively
mild despite the radiological changes, which can be quite dramatic because of the extensive calcifications of
ligamentous structures.5 Neurological complications due to cervical spinal cord compression have been
described and occur occasionally. 6, 35 DISH can predispose to chronic myelopathic symptoms. Reduced
flexibility and decreased range of spinal motion, as in ankylosing spondylitis, can result in fractures of the spine
even after minor trauma.11, 12, 36 These fractures tend to be unrecognized, unstable and associated with treatment
delays and permanent neurologic deficits.36, 12 There are two main reasons why delay in diagnosis often occurs:
on one hand the patient usually has a baseline level of spinal pain preventing him/her to seek medical attention in
case of altered pain pattern and on the other hand the treating physician doesn‘t suspect a spinal fracture because
the injury may have been relatively trivial.12 The findings of a fracture on plain radiographs can be subtle, so
additional CT-scans and/or MRI may be indicated to confirm the fracture and/or associated level of spinal cord
injury. Especially great care must be exercised when there is a history of trauma and neck pain. 12, 26, 35, 36 The risk
of spinal fractures may be increased in some advanced cases of DISH in which completely fused segments
exist.37 Symptoms of mechanical dysphagia, dyspnoea, stridor, thoracic outlet syndrome and heterotopic
ossification after hip arthroplasty are also mentioned 11. Patients can even die as a result of mechanical
respiratory failure, probably due to paralysis of the respiratory muscles. 11, 26, 36, 37, 38
According to Sreedharan and Li (2005) age may be a confounder in the relationship between DISH and the
increased incidence of spinal cord injuries, because other factors associated with increasing age, such as failing
eyesight and decreased mobility, could predispose to a growing risk of falling. When DM and DISH concur in
humans concomitantly, this could also make patients more susceptible to falls because of possible complications
of DM such as peripheral neuropathy, autonomic neuropathy, retinopathy, cataracts and episodes of
Treatment of DISH is usually non surgical but sometimes surgical intervention is indicated in case of specific
sequelae, such as fractures or dysphagia.11, 12 Nonsurgical therapy consisted of activity modification, physical
therapy, weight loss, corset or brace wear and medical therapy with nonsteroidal anti-inflammatory drugs
(NSAIDs).6, 39, 40 The efficacy of these therapies is still not well established, more research is necessary.12, 39 Pain
in the peripheral skeleton may respond to NSAIDs and analgesics. Pain from spurs can be treated with local
injection with lidocaine with or without steroids. If this does not help, surgical removal can be an option.6
II. DISH in dogs
2.1 Canine anatomy
The canine spine contains seven cervical, thirteen thoracic, seven lumbar and three (fused) sacral vertebral
bodies (Figure 4.).
Fig. 4. Canine skeleton.41
1. Wing of atlas, C1 vertebra; 2 axis, C2 vertebra; 5. C7
vertebra; T1. first thoracic vertebra; L1. first lumbar
vertebra; 13. sacrum
Figure 5, 6 and 7 show the cervical, thoracic and lumbar radiographs (lateral view) of the normal canine spine.
Fig. 5. Normal cervical radiograph of a dog.
Fig. 6. Normal thoracic radiograph of a dog.
Fig. 7. Normal lumbar radiograph of a dog (only part of L1 at left side of the radiograph).
The ventral longitudinal ligament is situated at the ventral aspect of the canine vertebral column (Figure 8).
Fig. 8. Ligaments of the vertebral colum.42
1. Supraspinous ligament; 2. spinous process;
3. interspinous ligament; 4. arch of vertebra;
5. interarcuate ligament; 6. intervertebral
foramen; 7. dorsal longitudinal ligament; 8.
ventral longitudinal ligament; 9.
2.2 DISH in dogs
For more than 40.000 years dogs have lived with humans in comparable conditions. They share the same
environment and are subjected to the same toxic noxae. Nowadays commercial dog foods are of high quality and
usually have a condensed form. The dog‘s nutrition is well balanced but a growing number of dogs become
obese.43 Dogs develop similar diseases as humans.44 Endocrinological syndromes and conditions of the spine
show great similarities between humans and dogs. Over the last 50 years the knowledge in veterinary medicine
has greatly increased and improved veterinary care for small animals has led to a higher live expectancy in dogs.
As in humans, the canine genome is available to the public.
In dogs diabetes mellitus is more a result of autoimmune mechanisms affecting the β-cell function and therefore
diabetes mellitus in dogs is more comparable to the human type I than type II. 44, 45
Surprisingly, only a few articles are published on DISH in animals. Hyperostosis is described in old Rhesus
monkeys.46 Only two case reports of DISH in dogs were published. 47, 48
In the case report of the 4-year-old Labrador with DISH Woodard et al (1985) described that the first
abnormalities were seen on the caudal proximal third of the right femur and appeared to extend caudally to the
ischium and cranially to the ilium. The authors stated that the changes in the right hip were associated with
spondylosis deformans of lumbar segments 6 and 7. Later more abnormal calcifications were noted in spinal and
numerous extra spinal locations. Woodard et al (1985) postulated that extraspinal alterations may be more
extensive than spinal lesions and can exist without vertebral abnormality and still be called DISH. When the
alterations in the lumbar region progressed the authors claimed that the lesions are similar to those in spondylosis
deformans but more exuberant. In addition, they claimed that other osseous lesions in this dog were distinctly
different; making this a dog with DISH and not just with spondylosis deformans. Further, Woodard et al (1985)
stated that there are certain aspects of canine DISH that have common features with lesions like spondylosis
deformans and spinal osteoartrosis. But these may have developed because of increased mobility of adjacent
segments of the spine and often they are associated with a prolapsed intervertebral disc. 47
Morgan and Stavenborn (1991) described a 4-year-old female Great Dane Dog with heavy new bone formations
throughout the thoracic and lumbar spine dorsally as well as at the lumbar and sacral regions dorsally and
ventrally, resulting in fusion of vertebral segments. This dog also demonstrated extra spinal new bone
abnormalities at several locations. They claimed that the radiographic and pathologic features of DISH in the dog
may closely resemble extensive spondylosis deformans but possess marked radiographic and morphologic
differences. For example vertebral osteophytes associated with spondylosis deformans typically center on
individually degenerated discs and do not have patterns of flowing bone growth involving contiguous segments
or dorsal periarticular changes.48 This is consistent with what Haller et al (1989) reported in humans. 18 Morgan
and Stavenborn (1991) brought forward the possibility that DISH in dogs had been described as a variant of
2.3 Spinal skeleton of a dog with lumbar DISH
Figure 9, 10 and 11 show photographs of a canine spine with macroscopic abnormalities resembling DISH,
found at the ventral side of the thoracic and lumbar region of the spine.
Fig. 9. Macroscopic photograph of a canine spine with DISH.
Fig. 10 and 11. Ventral view of the lumbar region of a canine spine with DISH. Note the flowing pattern of new
In the thoracic region, although not that extensive as in the lumbar region, bony outgrowths at the ventral part of
the spine are visible. But on the radiographs of this particular spine the spurs in the thoracic region do not seem
to make contiguous bridges (Figure 12).
Fig. 12. Thoracic region of a spine of a dog with abnormal bone growth.
On the other hand, in the lumbar region flowing contiguous calcification at the ventral part of the spine is
distinct. Some bridges may appear not completely continuous but enough closed segments remain to classify this
as DISH (Figure 13).
Fig. 13. Lumbar region of a spine of a dog with DISH.
2.4 CT and MRI of a dog with DISH
A 10-year-old Border Collie underwent diagnostic imaging for low back pain. On CT and MRI scan extensive
ventral bone formation was found in the lumbar region from L3 to L7 (Figure 14, 15 and 16). On T2-weighted
MRI scans a normal water signal (white) was found between L3 and L7, indicating normal disci at levels with
DISH. Caudally, at the level of L7-S1, were no bridge was visible on the MRI and the water signal was absent,
evidence of a hernia nuclei pulposi was found (Figure 16).
Fig. 14. CT scan of a canine vertebral body with ventral bony outgrowth.
Fig. 15. Transverse CT scan of a canine vertebral body with ventral bony outgrowth. Note the original ventral
lining of the vertebral body (^^).
Fig. 16. T2 weighted MRI of a canine lumbar spine with DISH. Note the intact normal nucleus pulposus water
signal (white) at levels L3 to L7 with DISH.
III. DISH versus spondylosis deformans
3.1 Spondylosis deformans in humans
In humans, spondylosis deformans is thought to originate from abnormalities in the peripheral fibers of the
annulus fibrosus. This will lead to weakening of the anchorage of the intervertebral disc to the vertebral body
and can result in displacement of the disc. This might lead to traction at the site of osseous attachment of the
annulus fibrosus to the vertebral surface were osteophytes can develop. 49 Osteophytes grow first in ventral and
then in cranial or caudal direction, this in contrast to the osteophytes in ankylosing spondylitis which grow only
cranial or caudally, forming the bamboo like spine. 49 Thereby in humans, the osteophytes in spondylosis
deformans probably originate from the annulus fibrosus and ventral body while in spinal DISH they grow on the
longitudinal ligament. Still the pathologic and radiographic features of both diseases are, in part, identical and
the precise relationship is not clear.18, 49 Like in DISH, prevalence of spondylosis deformans in humans increases
with age and is more seen in males and in patients with occupations which require heavy physical labour. 49
Haller et al (1989) investigated human pelvis and tried to differentiate between spondylosis deformans and DISH
on the basis of the pattern and extent of spinal ossification. In this study spondylosis deformans was diagnosed
on the basis of focal osteophytes and the absence of contiguous or flowing calcification or ossification and
patients were subdivided on number and size of the not contiguous osteophytes.18 This is not consistent to the
method Langeland and Lingaas (1995) used to determine degrees of spondylosis in Boxers since in that
subdivision also contiguous bridges could still be called spondylosis deformans. 9 Although Haller et al (1989)
described that in the 1960‘s and 1970‘s some authors considered DISH to be a form of spondylosis, Haller et al
postulated that DISH should be considered as a distinct entity which differs from spondylosis deformans not
only by the contiguous aspect of the ossification, but also by the dominance of ligamentous ossification in the
spine and also in extraspinal locations.18 Kiss et al (2002) compared a group of DISH patients with a group of
patients with spondylosis. They found differences in BMI, the occurrence of DM and the serum level of uric
acid. This higher level of uric acid was not associated with BMI, suggesting that obesity is not the cause for
those elevated levels.15 Other authors, strongly postulated that DISH is a distinct disorder with factors that
distinguish it from other vertebral diseases in humans.15 ,16, 17
3.2 Spondylosis deformans in dogs
Spondylosis deformans in dogs is often described in a way comparable to DISH. Instead of ossification of the
ligament, spondylosis deformans is supposed to be a degenerative disease of the spinal column characterised by
osteophytes forming a spur on the vertebral bodies originating from the end-plate region. Also, like in DISH, the
etiology of spondylosis deformans is not clear.9, 50 In contrast with DISH, sclerosis and vacuum phenomenon can
be found in canine spondylosis deformans. In dogs, both the narrowing of the intervertebral disc space and the
sclerotic appearance of the vertebral end-plate were common secondary to fractures, dislocations,
discospondylitis, malignant tumor or osteomyelitis. Without those latter changes the disc space is usually of
normal width.51 Severe spondylosis in dogs causes stiffness, lameness, change of gait and pain. 52 It can be
subdivided into 3 subclasses according to the degree of osteophytes development. In spondylosis (degree 3) the
osteophytes extend beyond the edge of the end-plate forming radiographically bony bridges between vertebrae. 9,
Three bridges in a row in spondyslosis (degree 3) may radiographically resemble DISH. In Boxers in Italy and
Norway prevalences and estimates of heritability for spondylosis deformans were reported. In Norway a
prevalence of 26% (104/402) for spondylosis deformans in Boxers over 1 year old was found. 9 In Italy a
prevalence of 50% for degree 3 was found in Boxers of one year and older. 52 The prevalence and the degree of
spondylosis increased with age and were high when all degrees were combined. 50, 52 Very high frequencies for
sites within the last three thoracic and first lumbar vertebrae were found.50, 52 Also a marked increase was found
for the site between the last lumbar and first sacral vertebra. 52 It is not easy to distinguish between DISH and
severe spondylosis deformans on a radiograph (Figure 12, 13 and 17). Morgan and Stavenborn (1991) brought
forward the possibility that DISH in dogs had been described as a variant of spondylosis deformans. 48
Figure 17 shows a radiograph of a dog diagnosed with spondylosis deformans. The contiguous bridges of L4-L7
could also be diagnosed as DISH. The bony abnormalities of T13-L4 show a more resemblance to the changes
compatible with spondylosis deformans.
Fig. 17. Radiograph of a dog diagnosed with spondylosis deformans. Note the variable pattern of osteophytes
ranging from small and interlocking finger like projections to massive ankylosing bridges and how the bridges of
L4-L7 resemble the appearance of DISH 53.
Wright (1982) made in her articles on canine vertebral osteophytes a distinction between end-plate osteophytes
(Type 1) compatible with those in spondylosis deformans and three other types (Types 2, 3 and 4) comparable
with those seen in human ankylosing hyperostosis (former name for DISH). The latter three formed spurs with a
broader base of origin at the vertebral body up to a contiguous ventral band of new bone (Figure 18) suggesting
that ankylosing hyperostosis can occur in dogs.54, 55
Fig. 18. Types of canine osteophytes according to Wright (1982). 55
IV. The aim of the study
The main objective of this study was to determine the prevalence of diffuse idiopathic skeletal hyperostosis
(DISH) in an out-patient population of dogs over 1 year of age. Secondary fields of interest were the possible
contribution of age, gender and breed differences to DISH as well as its association with obesitas or endocrine
diseases such as diabetes mellitus.
Materials and Methods
Radiographs were selected from dogs over 1 year of age that were referred to the University Clinic for
Companion Animals between February 2003 and January 2008. All patients were referred to this clinic by
general practitioners for various medical conditions. Radiographs were only used when more than 20 dogs of the
same breed were available, leading to the inclusion of 33 breeds. Radiographs of the thorax, abdomen and the
spine (cervical and thoracolumbar) were used for screening.
From this group, consisting of 2139 patients, the available radiographs were pre-screened by one investigator
(H.C. Kranenburg), according to a predefined scoringsystem: 1) Definitely DISH, 2) possibly DISH or other
pathology, 3) definitely not DISH or other pathology. Radiographs that were marked with the first two scorings
were re-reviewed by two investigators (L.A. Westerveld and H.C. Kranenburg) in a consensus meeting to
confirm the diagnosis DISH. In case no consensus on the diagnosis could be established a third investigator (J.J.
Verlaan) was consulted. In every radiograph the region of the spine that allowed screening for DISH was
identified and recorded. For example in some cases only a cervical region was available for evaluation and in
other cases only a thoracic and/or lumbar region could be reviewed. If only a part of such region or just a part of
the vertebral bodies was on the radiograph, the radiograph was excluded. If the facet joints or the disc spaces
could not be properly assessed on a radiograph, the radiograph was excluded.
Radiographs were all taken in the Division of Diagnostic Imaging using the standard methods for cervical,
thorax, abdomen and vertebral radiographs in at least ventro-dorsal and lateral projection.
Of all patients the identification number, gender, breed, date of birth and date of radiograph were recorded.
All abnormal features of the spine and its location were recorded. DISH was confirmed when the Resnick
criteria , as given at page 3, were met on the ventro-dorsal or lateral views or both. Ossification of the ventral
longitudinal ligament over three contiguous vertebral bodies (i.e. bridging of two disc spaces) was considered a
precursor of full blown DISH. This was defined as pre-stage DISH and recorded separately.23
The radiographic reports of the DISH patients made by radiologists were examined in order to get an impression
if and how the radiological features were noted.
The medical records of the patients with DISH and pre-stage DISH were examined in order to extract clinical
data from the medical history. The body weight of all patients with DISH was recorded from the medical chart
as well as the indication for the radiographic examination.
Statistical analysis was performed using SPSS software (version 16.0). Logistic regression analysis was
performed to investigate the influence of age, breed, gender and neutering status on the prevalence of DISH. To
differentiate correctly individuals diagnosed with pre-stage DISH were regarded as non DISH subjects. P values
less than 0.05 were considered significant.
The study started out with a cohort of 2139 patients. During the study 98 patients were excluded for a variety of
reasons. Patients were excluded because the quality of the radiographs was poor, others because the field of
interest was not completely visible on the radiograph and some were excluded because the date of birth was not
recorded. Of the group of 2041 patients 51.3% were male and 48.7% female dogs (Table 1). Of the male dogs
61.4% were neutered, whereas 31.6% of the female dogs were neutered. The mean age of all dogs was 7.2 years
(range 1-17 years), the mean age of those with DISH was 8.9 years (range 2-16 years), of the patients with pre-
stage DISH 8.5 years (range 2-14 years) and of the group with no DISH 7.0 years (range 1-17 years).
Table 1. Gender distribution.
Male neutered 643
Male intact 404
Male total 1047
Female neutered 314
Female intact 680
Female total 994
The overall frequency of DISH in the cohort of 2041 patients which were 1 year and older, was 3.8% (95%
confidence interval (CI) 3.0-4.7) and of pre-stage DISH 1.5% (95% CI 1.0-2.0). In the male dogs, the frequency
of DISH was 4.3% (95% CI 3.1-5.5) and of pre-stage DISH 1.7% (95 CI 0.9-2.5). In female dogs this was 3.3%
for DISH (95% CI 2.2-4.4) and 1.2% for pre-stage DISH (95% CI 0.5-1.9) (Table 2). The difference between the
incidence of DISH in male and female dogs over 1 year was not significant.
Table 2. Frequency of DISH and pre-stage DISH by gender: number of patients (percentage; 95% confidence
Male dogs Female dogs Total
No DISH 984 (94; 92.6-95.4) 949 (95.5; 94.2-96.8) 1933 (94.7; 93.74-95.68)
Pre-stage DISH 18 (1.7; 0.9-2.5) 12 (1.2; 0.5-1.9) 30 (1.5; 1.0-2.0)
DISH 45 (4.3; 3.1-5.5) 33 (3.3; 2.2-4.4) 78 (3.8; 3.0-4.7)
Total 1047 (100.0) 994 (100.0) 2041 (100.0)
The frequency was slightly higher in neutered dogs than in intact ones (Table 3) but no statistically significant
difference was found.
Table 3. Frequency of DISH and pre-stage DISH in intact and neutered dogs: number of patients (percentage;
95% confidence interval).
M intact M* F intact F*
No DISH 609 (94.7; 93.0-96.4) 375 (92.8; 90.3-95.3) 302 (96.2; 94.1-98.3) 647 (95.1; 93.5-96.8)
Pre-stage DISH 9 (1.4; 0.5-2.3) 9 (2.2; 0.8-3.7) 2 (0.6; -0.24-1.52) 10 (1.5; 0.6-2.4)
DISH 25 (3.9; 2.4-5.4) 20 (5.0; 2.8-7.1) 10 (3.2; 1.2-5.1) 23 (3.4; 2.0-4.7)
Total 643 (100.0) 404 (100.0) 314 (100.0) 680 (100.0)
(M = male dogs, F = female dogs, * = neutered)
The frequencies of DISH and pre-stage DISH increased with age (Table 4).
Table 4. The frequency of DISH and pre-stage DISH according to age groups: number of patients (percentage;
95% confidence interval).
Age (years) No DISH DISH Pre-stage DISH Total (n)
1 152 (100.0) 0 (0.0) 0 (0.0) 152
2 129 (98.5; 96.4-100.6) 1 (0.8; -0.7-2.3) 1 (0.8; -0.7-2.3) 131
3 140 (96.6; 93.6-99.5) 4 (2.8; 0.1-5.4) 1 (0.7; -0.7-2.0) 145
4 144 (97.3; 94.7-99.9) 2 (1.4; -0.5-3.2) 2 (1.4; -0.5-3.2) 148
5 165 (95.4; 92.3-98.5) 6 (3.5; 0.7-6.2) 2 (1.2; -0.4-2.8) 173
6 197 (97.0; 94.7-99.4) 3 (1.5; -0.2-3.1) 3 (1.5; -0.2-3.1) 203
7 197 (97.5; 95.4-99.7) 4 (2.0; 0.1-3.9) 1 (0.5; -0.5-1.5) 202
8 210 (94.6; 91.6-97.6) 10 (4.5; 1.8-7.2) 2 (0.9; -0.3-2.1) 222
9 172 (90.5; 86.4-94.7) 13 (6.8; 3.3-10.4) 5 (2.6; 0.4-4.9) 190
10 159 (90.3; 86.0-94.7) 12 (6.8; 3.1-10.5) 5 (2.8; 0.4-5.3) 176
11 116 (90.6; 85.6-95.7) 8 (6.3; 2.0-10.4) 4 (3.1; 0.1-6.1) 128
12 82 (90.1; 84.0-96.2) 7 (7.7; 2.2-13.2) 2 (2.2; -0.8-5.2) 91
13 45 (49.5; 39.2-59.7) 6 (11.5; 2.9-20.2) 1 (1.9; -1.8-5.7) 52
14 17 (89.5; 75.7-103.3) 1 (5.3; -4.8-15.3) 1 (5.3; -4.8-15.3) 19
15 4 (100.0) 0 (0.0) 0 (0.0) 4
16 3 (75; 32.6-117.4) 1 (25; -17.4-67.4) 0 (0.0) 4
17 1 (100.0) 0 (0.0) 0 (0.0) 1
Total 1933 (94.7; 93.7-95.7) 78 (3.8; 3.0-4.7) 30 (1.5; 1.0-2.0) 2041
The frequency of DISH in dogs age over 7 years was 5.7% (95% CI 4.3-7.1). In male dogs this was 7.1% (95%
CI 4.9-9.2) and in female dogs 4.4% (95% CI 2.7-6.1), which resulted in a male-female ratio of 1.6 to 1. The
frequency of pre stage DISH in dogs age 7 years and older was 1.9% (95% CI 1.1-2.8). In male dogs this was
2.6% (95% CI 1.3-4.0) and in female dogs 1.3% (95% CI 0.3-2.2) (Table 5).
Table 5. The frequency of DISH and pre-stage DISH in dogs age 7 years and older: number of patients
(percentage; 95% confidence interval).
Male dogs Female dogs Total
No DISH 487 (90.4; 87.9-92.8) 519 (94.4; 92.4-96.3) 1006 (92.4; 90.8-94.0)
Pre-stage DISH 14 (2.6; 1.3-4.0) 7 (1.3; 0.3-2.2) 21 (1.9; 1.1-2.8)
DISH 38 (7.1; 4.9-9.2) 24 (4.4; 2.7-6.1) 62 (5.7; 4.3-7.1)
Total 539 (100.0) 550 (100.0) 1089 (100.0)
Logistic regression analysis showed that age was significantly related to the presence of DISH. The odds ratio
was 1.260 (95% CI 1.169-1.358; P=0.000). In other words when age increased the risk for developing DISH
increased as well. DISH and pre-stage DISH were also found at relatively young age; a 2 year old dog was found
with DISH and an other 2 year old with pre-stage DISH (Table 4).
In 6 dogs no body weight was available. The body weight of neutered dogs seemed not higher than that of the
dogs that were not neutered (Table 6). The body weight of the dogs with DISH ranged from 8.7 kg for a 12-
year-old male West Highland White Terriër to 76 kg for a 7-year-old male Newfoundlander with a median of 38
kg. In dogs with pre-stage DISH the body weight ranged from 12 kg to 53.3 kg with a median of 33.5 kg. In
three records of patients with pre-stage DISH and in five records of patients with DISH the dog had been marked
Table 6. Number of dogs with DISH according to body weight classes and gender.
Body weight (kg) Total Male Female M intact M* F intact F*
<5 0 0 0 0 0 0 0
5—10 1 1 0 1 0 0 0
10—15 0 0 0 0 0 0 0
15—20 0 0 0 0 0 0 0
20—25 1 0 1 0 0 0 1
25—30 17 4 13 2 2 4 9
30—35 19 10 9 7 3 1 8
35—40 15 12 3 6 6 0 3
40—45 6 5 1 2 3 1 0
45—50 7 7 0 3 4 0 0
50—55 3 1 2 0 1 1 1
55—60 4 3 1 2 1 1 0
60—65 0 0 0 0 0 0 0
65—70 0 0 0 0 0 0 0
70—75 0 0 0 0 0 0 0
75—80 1 1 0 1 0 0 0
not known 4 1 3 1 0 2 1
Total n 78 45 33 25 20 10 23
Average weight 36 39 32 38 39 36 31
( M = male dogs, F = female dogs, * = neutered)
Figure 19, 20, 23 and 24 show the cervical, thoracic, lumbar and thoracic-lumbar radiographs (lateral view) of
dogs with DISH.
Fig. 19. Cervical radiograph of a 12 year old neutered male Rottweiler with DISH at C2-C5.
Fig. 20. Thoracic radiograph of a 10 year old female Boxer with DISH at T1-T12.
Logistic regression analysis showed that breed was significantly related to the presence of DISH. The odds ratio
was 0.942 (95% CI 0.918-0.967; p=0.000).
Table 9. DISH per breed: number of patients (percentage; 95% confidence interval).
Breed: No DISH DISH Pre-stage DISH ras/ ras/ Total
(%) total (%) (n)
Beagle 41 (100.0) 0 (0.0) 0 (0.0) 0.0 0.0 41
Belgian Shepherd, Malinois 51 (96.2; 91.1-101.4) 1 (1.9; -1.8-5.6) 1 (1.9; -1.8-5.6) 0.5 1.3 53
Bernese Mountaindog 193 (97.0; 94.6-99.4) 4 (2.0; 0.1-4.0) 2 (1.0; -0.4-2.4) 0.5 0.7 199
Dogue de Bordeaux 34 (100.0) 0 (0.0) 0 (0.0) 0.0 0.0 34
Border Collie 25 (96.2; 88.8-103.6) 1 (3.9; -3.6-11.3) 0 (0.0) 1.0 0.0 26
Bouvier des Flandres 59 (93.7; 87.6-99.7) 4 (6.4; 0.3-12.4) 0 (0.0) 1.7 0.0 63
Boxer 38 (55.1; 43.3-66.8) 28 (40.6; 29.0-52.2) 3 (4.4; -0.5-9.2) 10.7 2.9 69
Bullmastiff 12 (100.0) 0 (0.0) 0 (0.0) 0.0 0.0 12
Cairn Terrier 33 (100.0) 0 (0.0) 0 (0.0) 0.0 0.0 33
Dachshund 85 (98.8; 96.6-101.1) 0 (0.0) 1 (1.2; -1.1-3.4) 0.0 0.8 86
Doberman Pincher 31 (86.1; 74.8-97.4) 0 (0.0) 5 (13.9; 2.6-25.2) 0.0 9.3 36
Dutch Patridge Dog 18 (94.7; 84.7-104.8) 0 (0.0) 1 (5.3; -4.8-15.3) 0.0 3.5 19
Great Dane 29 (96.7; 90.3-103.1) 0 (0.0) 1 (3.3; -3.1-9.8) 0.0 2.2 30
German Shepherd 100 (88.5; 82.6-94.4) 8 (7.1; 2.4-11.8) 5 (4.4; 0.6-8.2) 1.9 2.9 113
German Pointer 35 (92.1; 83.5-100.7) 3 (7.9; -0.7-16.5) 0 (0.0) 2.1 0.0 38
English Bulldog 28 (100.0) 0 (0.0) 0 (0.0) 0.0 0.0 28
English Cocker Spaniel 43 (100.0) 0 (0.0) 0 (0.0) 0.0 0.0 43
Flatcoated Retriever 51 (83.6; 74.3-92.9) 8 (13.1; 4.6-21.6) 2 (3.3; -1.2-7.8) 3.5 2.2 61
French Bulldog 61 (100.0) 0 (0.0) 0 (0.0) 0.0 0.0 61
Golden Retriever 150 (96.2; 93.1-99.2) 6 (3.9; 0.8-6.9) 0 (0.0) 1.0 0.0 156
Irish Setter 21 (95.5; 86.7-104.1) 1 (4.6; -4.2-13.3) 0 (0.0) 1.2 0.0 22
Jack Russell Terrier 163 (100.0) 0 (0.0) 0 (0.0) 0.0 0.0 163
Labrador Retriever 233 (95.5; 92.9-98.1) 6 (2.5; 0.5-4.4) 5 (2.0; -.3-3.8) 0.6 1.4 244
Leonberger 21 (84.0; 69.6-98.4) 3 (12.0; -0.7-24.7) 1 (4.0; -3.7-11.7) 3.2 2.7 25
Maltese 71 (100.0) 0 (0.0) 0 (0.0) 0.0 0.0 71
Newfoundland 20 (90.9; 78.9-102.9) 2 (9.1; -2.9-21.1) 0 (0.0) 2.4 0.0 22
Poodle 25 (100.0) 0 (0.0) 0 (0.0) 0.0 0.0 25
Rhodesian Ridgeback 26 (96.3; 89.2-103.4) 0 (0.0) 1 (3.7; -3.4-10.8) 0.0 2.5 27
Rottweiler 117 (97.5; 94.7-100.3) 2 (1.7; -0.6-4.0) 1 (0.8; -0.8-2.5) 0.4 0.6 120
Staffordshire Bull Terrier 20 (100.0) 0 (0.0) 0 (0.0) 0.0 0.0 20
Weimaraner 22 (100.0) 0 (0.0) 0 (0.0) 0.0 0.0 22
West Highland White Terrier 40 (95.2; 88.8-101.7) 1 (2.4; -2.2-7.0) 1 (2.4; -2.2-7.0) 0.6 1.6 42
Yorkshire Terrier 37 (100.0) 0 (0.0) 0 (0.0) 0.0 0.0 37
Total 1933 (94.7; 93.7-95.7) 78 (3.8; 3.0-4.7) 30 (1.5; 1.0-2.0) 2041
Of the 33 breeds, the following 13 breeds had no DISH or pre-stage DISH:
Dogue de Bordeaux
English Cocker Spaniel
Jack Russell Terrier
These 13 breeds amounted to a total of 592 dogs and included almost all small breeds. In the other small breeds
(i) only one West Highland White Terrier had DISH; (ii) one West Highland White Terrier had pre-stage DISH;
(iii) one Border Collie had DISH; and (iv) one Dachshund had pre-stage DISH. Thereby, DISH was not found in
any Rhodesian Ridgebacks, Dachshunds and Great Dane, whereas only one dog from each of these breeds was
diagnosed with pre-stage DISH. DISH, but no pre-stage DISH was found in the Border Collie, Bouvier des
Flandres, Golden Retriever and Newfoundlander breeds (Table 7). The frequency of DISH was particularly high
in Boxers and Flatcoated Retrievers: if these two breeds were left out an overall frequency of 2.2% (42/1918)
would remain. Two breeds, the Boxer and Flatcoated Retriever, were most commonly affected by DISH (Table
7). The breed that was most affected by DISH is the Boxer with a frequency of 56.3% in Boxers over 7 years
Table 8. Frequency of DISH in Boxers over 7 years: number of patients (percentage; 95% confidence interval).
Boxers > 7 years
No DISH 14 (43.8; 26.6-60.9)
DISH 18 (56.3; 39.0-73.4)
Pre-stage DISH 0 (0.0)
Total 32 (100.0)
In the patients with DISH the exact location (cervical, thoracic, and lumbar region) was identified on the
radiographs. The number and location of the individual bridges were marked (Table 9 and Figure 21).
Table 9. Frequency of DISH and pre-stage DISH according to location in the spine in 2041 dogs.
no DISH Pre-stage DISH Pre-stage DISH (%) DISH DISH (%)
Cervical 79 0 0,0 2 2,5
Cervical + Thoracic 133 0 0,0 0 0,0
Cervical + Lumbar 2 0 0,0 0 0,0
Thoracic 1407 20 1,4 52 3,5
Thoracic + Lumbar 144 3 1,8 22 13,0
Lumbar 132 7 5,0 2 1,4
Cervical + Thoracic + Lumbar 36 0 0,0 0 0,0
Total 1933 30 78
Two patients with DISH in the thoracic region also demonstrated pre-stage DISH in the cervical region. Two
others had DISH in the cervical region (Figure 19). Most of the patients with DISH or pre stage DISH had
bridges in the thoracic region (Table 8 and Figure 21).
When the number of bridges were counted at the segments which contributed to the classification of contiguous
calcification along four or more vertebrae, especially the second half of the thoracic vertebral bodies and the first
half of the lumbar vertebrae were involved (Figure 21 and 22).
Bridges in 78 dogs with DISH
Number of bridges
c1- c2- c3- c4- c5- c6- c7- t1- t2- t3- t4- t5- t6- t7- t8- t9- t10- t11- t12- t13- L1- L2- L3- L4- L5- L6- L7-
c2 c3 c4 c5 c6 c7 t1 t2 t3 t4 t5 t6 t7 t8 t9 t10 t11 t12 t13 L1 L2 L3 L4 L5 L6 L7 S1
Fig. 21. Distribution of bridges along the spine in 78 dogs with DISH.
Bridges in 28 Boxers with DISH
Number of bridges
c1- c2- c3- c4- c5- c6- c7- t1- t2- t3- t4- t5- t6- t7- t8- t9- t10- t11- t12- t13- L1- L2- L3- L4- L5- L6- L7-
c2 c3 c4 c5 c6 c7 t1 t2 t3 t4 t5 t6 t7 t8 t9 t10 t11 t12 t13 L1 L2 L3 L4 L5 L6 L7 S1
Fig. 22. Distribution of bridges along the spine in 28 Boxers with DISH.
The retrospective analysis of the medical records showed that none of the patients were diagnosed with Diabetes
Mellitus type I or II. Furthermore no similarities in the medical background were found in the medical records of
the patients with either DISH or pre-stage DISH. The indications for the radiographs were reviewed in the
records and classified into five groups (Table 10).
Table 10. Indications for radiographic examination.
Radiograph Indication DISH Pre-stage DISH
Thorax - metastases 43 15
Thorax – other 29 10
Abdomen 7 4
Spine 17 6
In some reports multiple indications for the radiographic examination of those particular patients had been
recorded. The group thorax-other included patients with either suspicion of respiratory, cardiac or esophageal
In all patients with DISH or pre-stage DISH, the report of the radiologist was available for evaluation. In 82.0%
(64/78) of the DISH patients and 70.0% (21/30) of the pre-stage DISH patients the radiologist made a remark of
either (severe) spondylosis deformans or (severe) ossification of the axial skeleton.
In two cases (2.6%) DISH was mentioned. In all of the reports the vertebral abnormalities of spondylosis,
ossification or DISH was considered an accidental finding and not related to the complaints or symptoms.
Figure 23 and 24 show radiographs of two Boxers with DISH. The bridges are contiguous ventral of the
Fig. 23. Lumbar radiograph of an 8 year old neutered male Boxer with DISH at L4-S1.
Fig.24. Radiograph of an 8 year old neutered female Boxer with DISH at T7-L7.
The bony abnormalities in the spine of patients with DISH originate from the ventral longitudinal ligament and
in patients with spondylosis deformans from the vertebral bodies end plate; however, in some cases
differentiation between the two disorders may be difficult. 8, 48, 49
To the best of our knowledge, no further studies of DISH in dogs were reported besides the two case reports of
Morgan and Stavenborn (1991) and Woodard et al (1985). However, Morgan and Stavenborn (1991) postulated
that DISH in dogs is a rather common disease of large and giant breeds at a relative young age. These authors
suggest seven criteria for diagnosing DISH, including flowing ossification along three instead of four vertebral
bodies and the absence of localized spondylosis deformans. 48 It is not clear why they thought it was necessary to
change the Resnick (1976) criteria. Both dogs were eventually euthanized because of extreme stiffness, pain and
limitation in range motion of the spine and several joints of the legs. 47, 48 In the present study, in 23 of the 108
patients with DISH or pre-stage DISH one of the indications for radiographic examination was a spinal and
locomotion problem. In some cases concurrent abnormalities appeared to be present and in other cases, also in
those with advanced cancer, the patients were euthanized without a definitive diagnose for the locomotion
problems. It is possible that some dogs with DISH were without symptoms but on the other hand can it be likely
that other dogs with DISH suffered from pain and stiffness. Thereby, it is possible that when a dog has a fracture
of the spine and is paralyzed, no diagnostic research is done and the dog is euthanized by the general
veterinarian. When veterinarians become aware of the symptoms associated with DISH in humans it is possible
their eyes will be opened for (mild) symptoms in dogs with DISH. Perhaps that in some cases dogs with DISH
will have a better spinal and/or extra spinal range of motion when given NSAID‘s. But more research is needed
Resnick and Niwayama (1976) reported that in humans the thoracic ossifications were more frequent in the 7th to
10th thoracic vertebrae. In addition, they noted a lower incidence in the most cranial thoracic vertebrae.8 In the
present study of dogs comparable results were found. On the other hand some patients with DISH could have
been overlooked because radiographs of the complete axial skeleton were not available in each patient studied.
But, in humans the thoracic radiographs are the golden standard for screening for DISH and of the patients with
―no dish‖ in this study 89% (1726/1940) were evaluated by thoracic radiographs. Also in dogs thoracic
radiographs can serve as a golden standard since 23/30 cases of pre-stage DISH and 74/78 cases of DISH were
present in the thoracic region.
The out-patient population in the present study is not completely representative for the population of dogs in The
Netherlands. Dogs from 33 of the most common breeds were included in this investigation. It is possible that the
dogs visiting a tertiary line, university clinic are generally younger than the dogs visiting a general clinic. A lot
of owners will not travel far with a geriatric dog with, for example, advanced cancer. Since the prevalence of
DISH increased with age, this could lead to an underestimation of the prevalence of DISH in the population of
dogs in the Netherlands. The radiological changes found in the spine can be very extensive and are not
exclusively found in older dogs. But with an average age of 8.9 years, patients with DISH are older than the
average dog in this study.
In human studies the minimum age of the studied patients is usually 50 years.14, 19, 21, 23, 24 Human age of 50 years
is somewhere consisted to be compatible with dogs age 7 years. Therefore the prevalence of dogs over 7 years
old is assessed to facilitate comparison between both species. Because the life expectation is not the same for all
breeds, exact comparison is difficult.
Prevalences of 3.8% for DISH and 1.5% for pre-stage DISH in 2041 dogs older than 1 year and 5.7% for DISH
and 1.9% for pre-stage DISH in 1089 dogs over 7 years were found. No significant statistic difference between
male and female dogs over 1 year of age was found. In dogs over 7 years old a male-female ratio of 1.6 to 1 was
found. The overall prevalence is not as high as the outcome of the Dutch study in human patients over 50 years,
which showed a male frequency of 22.7% and 12.1% in females, with a male-female ratio of 1.6 to 1 and a
frequency of pre-stage DISH of 4.6%.23
Looking at the differences between the breeds several things became apparent. It is not a disease for smaller
breeds and some breeds, like the Boxer and the Flatcoated Retriever are typically more predisposed. There may
be a genetic susceptibility to DISH. This is supported by differences in prevalence for DISH in African Blacks,
Afro-American, Caucasian, Native American and Asian humans.13, 19, 20, 21
In humans the association between obesity and DISH is well described and acknowledged. It is possible that this
could be a risk factor in dogs as well. In dogs, there is no uniform technique available to easily define obesity. A
variety of techniques can be used, some more accurate than others. It is a fact that more and more dogs are
overweight, incidences between 22% and 34% are published. 43 In only three records of patients with pre stage
DISH and in five records of patients with DISH the dog was marked as obese. However, the body weights were
measured after the time of referral when the dogs suffered from chronic diseases such as cancer and probably
lost some weight as a result thereof. Also, the recorded body weight was the weight which was measured at the
last time the dog was in the clinic. This means that if a dog was first obese and later on lost weight, only the
outcome of the last visit and measurement was available for evaluation.
In this study the author relied on the diagnosis by the specialists responsible for the patients or on the referring
veterinarians. A relation with an underlying metabolic abnormality was not found, more research and
prospective study is required on that subject.
Summary and Conclusion
Diffuse idiopathic skeletal hyperostosis (DISH) is a common, systemic disorder of the axial and peripheral
skeleton in humans. The etiology of DISH in humans is unknown and the condition occurs in middle aged and
elderly patients. It results in ossification of soft tissues such as longitudinal spinal ligaments, joints, sites of
attachment of tendons, muscles and/or capsules to bone. In mean, DISH is mostly seen in the elderly and
demonstrates a male predominance.
Prevalences of 25% in males and 15% in females have been reported. In humans the association between obesity
and DISH is well described and acknowledged. There may be a genetic susceptibility to DISH in humans. This is
supported by differences in prevalence for DISH in Afro-Americans, Caucasian, Native Americans and Asians.
In humans the variety of possible complaints due to DISH is nowadays well described, this also applies to the
risks for fractures and severe neurological deficits.
The main objective of the present study was to determine the prevalence of diffuse idiopathic skeletal
hyperostosis (DISH) in an out-patient population of dogs over 1 year of age in The Netherlands. The incidence
of DISH and pre-stage DISH found in dogs over 1 year of age was 3.8% in male and 1.5% in female dogs. In
male dogs over 7 years a prevalence of 7.1% was found. DISH can be diagnosed in dogs at a relative young age
and the prevalence increased with age. In this study no connection with diabetes mellitus or obesity was found.
In dogs, Boxers are predisposed for DISH. Prevalence over 50% in Boxers age 7 years and older was found.
Although the bony abnormalities in the spine of patients with DISH originate from the ventral longitudinal
ligament and in patients with spondylosis deformans from the end plate of the vertebral bodies, it is not easy to
radiographically differentiate between these disorders. It is possible that radiographic changes to date found in
the spine of dogs were incorrectly diagnosed as spondylosis deformans but were in fact abnormalities compatible
with DISH. The results of this study support this, as the original radiologists reports almost in every case named
the condition that was now identified as DISH in the present study, spondylosis deformans. More research,
preferable prospective, to the prevalence, symptoms and etiology is necessary as well as to the therapeutic
For the general veterinarian practitioners as well as orthopaedics and neurologists it is important to be aware of
the fact that radiological changes compatible with DISH could be the reason for the dog‘s complaints, as well
predispose them to fractures of the spine.
The author would like to thank Dr. B.P. Meij, Prof. Dr. H.A.W. Hazewinkel, Drs. L.A. Westerveld, Prof. Dr. G.
Voorhout, Prof. Dr. W. Dhert and E. Petersen for all their time, inspiration help and efforts before and during
1. Forestier J, Rotes-Querol J. Senile ankylosing hyperostosis of the spine. Ann Rheum Dis 1950;9:321-
2. Rothschild BM. Diffuse idiopathic skeletal hyperostosis as reflected in the paleontologic record:
dinosaurs and early mammals. Semin Arthritis Rheum 1987;17:119-125
3. Bjorkengren AG, Sartoris DJ, Shermis S, Resnick D. Patterns of paravertebral ossification in the
prehistoric saber-toothed cat. AJR 1987;148:779-782
4. Chhem RK, Schmit P, Faure C. Did Ramses II really have ankylosing spondylitis? A reappraisal. Can
Assoc Radiol J 2004;55:211-217
5. Verlaan JJ, Oner FC, Maat GJR. Diffuse idiopathic skeletal hyperostosis in ancient clergyman. Eur
Spine J 2007;16:129-35
6. Utsinger PD. Diffuse idiopathic skeletal hyperostosis. Clin Rheum Dis. 1985;11:325-351
7. Resnick D, Shaul SR, Robins JM. Diffuse idiopathic skeletal hyperostosis (DISH): Forestier‘s disease
with extraspinal manifestations. Radiology 1975;115:513-524
8. Resnick D, Niwayama G. Radiographic and pathologic features of spinal involvement in diffuse
idiopathic skeletal hyperostosis (DISH). Radiology 1976;119:559-568
9. Langeland M, Lingaas F. Spondylosis deformans in the boxer: Estimates of heritability. J Small Anim
10. Fornasier VL, Littlejohn G, Urowitz MB, Keystone EC, Smythe HA. Spinal entheseal new bone
formation: the early changes of spinal diffuse idiopathic skeletal hyperostosis. J Rheum 1983;10:939-
11. Sreedharan S, Li YH. Diffuse idiopathic skeletal hyperostosis with cervical spinal cord injury. Ann
Acad Med Singapore 2005;34:257-261
12. Balanger TA, Rowe DE. Diffuse Idiopathic Skeletal Hyperostosis: Musculoskeletal Manifectations.
Journal of the American Academy of Othopaedic Surgeons 2001;9:258-267
13. Childs SG. Diffuse idiopathic skeletal hyperostosis, Forestier‘s disease. Orthopaedic Nursing
14. Kiss Cs, O‘Neill TW, Mituszova M, Szilagyi M, Poor Gy. The prevalence of diffuse idiopathic skeletal
hyperostosis in a population-based study in Hungary. Scand J Rheumatol 2002;31:226-229
15. Kiss C, Szilagyi M, Paksy A, Poor G. Risk factors for diffuse idiopathic skeletal hyperostosis: a case-
control study, Rheumatology (Oxford) 2002;41:27-30
16. Sarzi-Puttini P, Atzeni F. New devolopmenst in our understanding of DISH. Curr Opin Rheumatol
17. Mata S, Fortin PR, Fitzcharles MA, Starr MR, Joseph L, Watts CS, Gore B, Rosenberg E, Chhem RK,
Esdaile JM. A controlled study of diffuse idiopathic skeletal hyperostosis. Medicine 1997;76:104-117
18. Haller J, Resnick D, Miller CW, Schils JP, Kerr R, Bielecki D, Sartoris DJ, Gundry CR. Diffuse
idiopathic skeletal hyperostosis: diagnostic significance of radiographic abnormalities of the pelvis.
19. Weinfeld RM, Olson PN, Maki DD, Griffiths HJ. The prevalence of diffuse idiopathic skeletal
hyperostosis (DISH) in two large American Midwest metropolitan hospital populations. Skeletal Radiol
20. Cassim B, Mody GM, Rubin DL. The prevalence of diffuse idiopathic skeletal hyperostosis in African
blacks. Brit J Rheum 1990;29:131-132
21. Kim SK, Choi BR, Kim CG, Chung SH, Choe JY, Joo KB, Bae SC, Yoo DH, Jun JB. The prevalence
of idiopathic skeletal hyperostosis in Korea. J Rheumatology 2004;31:2032-2035
22. Mader R, Dubenski N, Lavi I. Morbidity and mortality of hospitalized patients with diffuse idiopathic
skeletal hyperostosi. Rheumatol Int 2005;26:132-136
23. Westerveld LA, Quarles van Ufford HME, Verlaan JJ, Oner FC. The prevalence of diffuse idiopathic
skeletal hyperostosis in an outpatient population in the Netherlands. J Rheumatology 2008;35:(in press)
24. Pappone N, Lubrano E, Espesito-del Puente A, d‘Angelo S, Di Girolamo C, Del Puente A. Prevalence
of diffuse idiopathic skeletal hyperostosis in a female Italian population. Clin Exp Rheumatol
25. Mata S, Hill RO, Joseph L, Kaplan P, Dussault R, Watts CS, Fitzcharles MA, Shiroky JB, Fortin PR,
Esdaile JM. Chest radiographs as a screening test for diffuse idiopathic skeletal hyperostosis. J
26. Meyer PR jr. Diffuse idiopathic skeletal hyperostosis in the cervical spine. Clin Orthop 1999;359:49-57
27. El Miedany YM, Wassif G, El Baddini M. Diffuse idiopathic skeletal hyperostosis (DISH): is it af
vascular aetiology? Clin Exp Rheumatol 2000;18:193-200
28. Hai Li, Lei-Sheng Jiang, Li-Yang Dai. Hormones and growth factors in the pathogenesis of spinal
ligament ossification. Eur Spine J 2007;1075-84
29. Sencan D, Elden H, Nacitarhan V, Sencan M, Kaptanoglu. The prevalence of diffuse idiopathic skeletal
hyperostosis in patients with diabetes mellitus. Rheumatol Int 2005;25:518-521
30. Vezyroglou G, Mitropoulos A, Kyriazis N, Antoniadis C. A metabolic syndrome in Diffuse Idiopathic
Skeletal Hyperostosis. A controlled study. J Rheumatology 1996;23:672-676
31. Denko CW, Boja B, Moskowitz RW. Growth promoting peptides in osteoarthritis and diffuse idiopathic
skeletal hyperostosis- Insulin, Insulin-like growth factor-I, growth hormone. J Rheumatol
32. Daragon A, Mejjad O, Czernichow P, Louvel JP, Vittecoq O, Durr A, Le Loet X. Vertebral
hyperostosis and diabetes mellitus, a case-control study. Annals of the Rheumatic diseases 1995;54:375-
33. Pappone N, Di Girolamo, Del Puente A, Scarpa R, Oriente P. Diffuse idiopathic skeletal hyperostosis
(DISH) a retrospective analysis. Clin Rheumatol 1996;15:121-124
34. Olivieri I, D‘Angelo S, Cutro MS, Padula A, Peruz G, Montaruli M, Scarano E, Giasi V, Palazzi C,
Khan MA. Diffuse idiopathic skeletal hyperostosis may give the typical postural abnormalities of
advanced ankylosing spondylitis. Rheumatology (Oxford) 2007;46:1709-11
35. Alenghat JP, Hallett M, Kido DK. Spinal cord compression in diffuse idiopathic skeletal hyperostosis.
36. Callahan EP, Aquillera H. Complications following minor trauma in a patient with diffuse idiopathic
skeletal hyperostosis. Ann of Emergency Medicine. 1993;2:1067-1070
37. Paley D, Schwartz M, Cooper P, Harris WR, Levine AM. Fractures of the spine in diffuse idiopathic
skeletal hyperostosis. Clin Orthop 1991;267:22-32
38. Julkunen H, Aromaa A, Knekt P. Diffuse idiopathic skeletal hyperostosis (DISH) and spondylosis
deformans as predictors of cardiovascular diseases and cancer. Scand J Rheumatology 1981;10:241-248
39. Troyanovich SJ, Buettner M. A structural chiropractic approach to the management of diffuse
idiopathic skeletal hyperostosis. Journal of manipulative and physiological therapeutics 2003;26:202-
40. Al-Herz A, Snip JP, Clark B, Esdaile JM. Exercise therapy for patients with diffuse idiopathic skeletal
hyperostosis. Clin Rheumatol 2008;27:207-210
41. Dyce KM, Sack WO, Wensing CJG. Veterinary Anatomy. Second edition. WB Saunders Company,
42. Dyce KM, Sack WO, Wensing CJG. Veterinary Anatomy. Second edition. WB Saunders Company,
43. German AJ. The growing problem of obesity in dogs and cats. J Nutr. 2006;136:S1940-1946
44. Rijnberk A, Kooistra HS, Mol JA. Endocrine diseases in dogs and cats, similarities and differences with
endocrine diseases in humans. Growth hormone & IGF research 2003;13:S158-S164
45. Catchpole B, Kennedy LJ, Davison LJ, Ollier WER. Canine diabetes mellitus: from phenotype to
genotype. J Small Anim Medicine 2008;49:4-10
46. Sokoloff L, Snell KC, Stewart HL. Spinal ankylosis. in old rhesus monkeys. Clin Orthop 1968;61:285-
47. Woodard JC, Poulos PW, Parker RB, Jackson RI, Eurell JC. Canine diffuse idiopathic skeletal
hyperostosis. Vet Pathol 1985;22:317-326
48. Morgan JP, Stavenborn M. Disseminated idiopathic skeletal hyperostosis (DISH) in a dog. Veterinary
49. Resnick R. Degenerative diseases of the vertebral column. Radiology 1985;156:3-14
50. Read RM, Smith RN, A comparison of spondylosis deformans in the English and Swedish cat and in
the English dog. J small Anim Pract 1968;9:159-166
51. Morgan JP. Spondylosis deformans in the dog: its radiolographic appearance. J Am Vet Radiol Soc
52. Carnier P, Gallo L, Sturaro E, Piccinini P, Bittante G. Prevalence of spondylosis deform ans and
estimates of genetic parameters for the degree of osteophytes development in the Italian Boxer dogs. J.
Anim. Sci 2004;82:85–92
53. Morgan JP, Biery DN. Chapter 61: Spondylosis deformans. In: Newton CD, Nunamaker DM. Textbook
of Small Animal Orthopaedics. JP Lipponcott Company. 1985.
54. Wright JA. A study of vertebral osteophytes formation in the canine spine. I. Spinal survey. J Small
Anim Pract 1982;23:697-711
55. Wright JA. A study of vertebral osteophytes formation in the canine spine. II. Radiograhic survey. J
Small Anim Pract 1982;23:747-761