Keywords abdomen - adolescent - prevalence - Scheuermann disease - scoliosis
Introduction
Scheuermann's disease (SD) was defined by Holger Werfel Scheuermann in 1920 and is
also known as osteochondritis juvenile dorsi or kyphosis dorsalis juvenilis.[1 ] Sorensen described the radiographic criteria for typical SD, including anterior
wedging greater than 5° in at least 3 adjacent vertebral bodies.[2 ]
Schmorl's nodes, irregular and flattened vertebral endplates, narrowed intervertebral
disc spaces, and anteroposterior elongation of the apical vertebral bodies are other
associated radiological features of SD.[2 ]
[3 ]
[4 ]
[5 ]
[6 ] The affection leads to rigid kyphosis of the median and lower thoracic, or upper
lumbar regions.[7 ]
Two patterns have been defined in SD based on the affected area of the vertebra.[7 ]
[8 ]
In the typical or classic form, the thoracic region is frequently affected and is
characterized by thoracic kyphosis increase and wedging of the vertebral bodies. This
pattern is also accompanied by nonstructural hyperlordosis of the cervical and lumbar
spine.[7 ]
[9 ]
[10 ]
[11 ] On the other hand, SD of atypical pattern (thoracolumbar or lumbar) has been defined
later and is distinguished from the typical one by lack of thoracic kyphosis and evident
wedging of the vertebrae.[4 ]
[12 ]
[13 ] Also, the apex of the kyphosis is located at the thoracolumbar junction (T11–T12)
in an atypical pattern of the disease.[4 ] Such pathologies account for 25 to 80% of SD.[4 ]
Despite the importance of SD, knowledge of its epidemiology needs to be improved.
Most articles mention that the prevalence varies from 0.4 to 8.0%,[2 ]
[7 ]
[14 ] and a few recent studies have examined the prevalence of SD, but only in individuals
aged from either 45 or 50 to 80 years or older, with a reported prevalence of 4.0
to 8.0%.[15 ] Besides, these studies are contradictory regarding the prevalence between the genders.
Some studies have suggested that the prevalence in men may be higher than in women,[3 ]
[8 ]
[16 ]
[17 ]
[18 ]
[19 ] while other publications suggest that it involves each gender equally.[15 ]
[20 ]
Furthermore, previous studies on the prevalence of SD did not evaluate the atypical
pattern of the condition separately.[7 ]
[15 ]
[20 ]
Thus, the present study aimed to determine the prevalence of atypical SD in patients
aged between 18 and 40 years using abdomen tomography as a screening tool.
Material and Methods
Study Design and Patient Population
This is a cross-sectional study. The hospital's Ethics Committee and Internal Review
Board approved the study protocol before initiation under number (CAAE: 51849821.6.0000.5440).
The study design was based on evaluating tomographic images of the abdomen. The exclusion
criteria used in the present study were patients aged < 18 years old or > 40 years
old, with previous spine surgery, presence of thoracolumbar spine diseases such as
fractures, tumors, infections, ankylosing spondylitis, and familial hyperostotic disease.
The inclusion criteria were patients aged between 18 and 40 years old and the presence
of adequate images for analysis according to previously established parameters.
Retrospectively, we selected abdomen tomographic images of 1,287 individuals (756
females and 531 males) obtained from a database of patients referred to a radiologic
private clinic from October 2021 to August 2022.
Computed Tomography (CT) Acquisition
Imaging exams were performed using a GE LightSpeed VCT device. The CT images obtained
were reconstructed with 1.25-mm thick slices for the axial plane. The reading and
reproduction of the images, as well as their reconstructions, were carried out using
the Carestream PACS software system, version 12.1.5.1156 (Carestream Health, Rochester,
NY, USA).
Image Analysis
The parameter to identify patients with atypical SD was wedging in 3 consecutive vertebrae
by 5° or more (measurement of the Cobb angle in each vertebra), with a total Cobb
angle of 10° or higher in the sagittal-plane tomographic images. The thoracolumbar
segment considered in our study was the one from T8 to L2 (thoracolumbar transition)
([Figs. 1 ]
[2 ]).
Fig. 1 Individual Cobb angle measurement in vertebra with wedging.
Fig. 2 Total Cobb angle measurement after identifying wedging greater than 5° in 3 consecutive
vertebrae (T10–T12).
As coronal image visualization was available in all the exams evaluated, in which
it was possible to observe the spine from the sacrum until the thoracic segment, it
was also possible to evaluate the presence or absence of scoliosis in patients. Therefore,
scoliosis was investigated using the CT coronal images. The method used for this was
the evaluation of the coronal alignment between the cervical, thoracic, and lumbar
vertebrae and the Cobb angle measurement ([Fig. 3 ]).
Fig. 3 Coronal slice used for diagnosing scoliosis in positive cases of atypical Scheuermann's
disease, as present in the above-mentioned case.
In cases classified as positive for atypical SD, the level of kyphosis and the presence
of Schmorl's nodules and scoliosis were also assessed. These patients were identified
according to sex and age to correlate the tomographic findings ([Fig. 4 ]). Data collection, through analysis of image exams, was carried out by a team composed
of 3 researchers using the Carestream PACS system software version 12.1.5.1156..
Fig. 4 Sagittal cut used for diagnosing Schmorl's nodes (at the level of L1) in positive
cases of the disease.
Statistical Analysis
The data were compiled into a Microsoft Excel software spreadsheet (Microsoft Corp.,
Redmond, WA, USA) and analyzed using the R software version 4.2 (R Foundation for
Statistical Computing, Vienna, Austria). Continuous data were expressed as mean, standard
deviation, median, and interquartile range (IQR), and categorical data were described
as absolute values and percentages.
The T-test or Yuen-Welch robust test was used to compare the mean between the two
groups, depending on the variable's distribution. Similarly, to compare three or more
groups of continuous data, the analysis of variance (ANOVA) test or the robust one-way
ANOVA test was used, with a trim of 0.2, depending on the distribution of the variable.
In situations of p < 0.05, multiple paired tests were performed with p -value adjustment using the Bejanmini-Yuketieli method, to verify which groups had
a significant difference.
The Shapiro-Francia test was used to test the distribution of variables. To compare
categorical variables, with only two groups, the Fisher or Chi-squared test was used,
according to the following rule: if more than 20% of the expected values of the counting
cells were greater than 5, the Fisher test was used, otherwise the Chi-squared test
was used (Kim, 2017).[26 ] In the case of categorical variables with more than 3 levels, the Chi-squared test
was used. Values of p < 0.05 were considered statistically significant.
Results
Tomographic images from 1,287 patients were randomly selected, with a median age of
33 (IQR: 5) years old (18–40 years old), 756 (58.8%) male and 531 (41.2%) female patients.
The median age among the men was 33 (IQR: 6) years old (19–40 years old), and among
the women, it was 33 (IQR: 10) years old (19–40 years old). There was no excluded
patient. Out of the 1,287 CT scans evaluated, 28 patients were diagnosed with atypical
SD, representing 2.2% of the total patients included in this study.
Age
Regarding patient age, younger patients (18–29 years old) presented higher Cobb values
than older patients (30–40 years old), with an average of 17.4° and 15°, respectively
([Fig. 5 ]). However, when evaluated by the Spearman correlation test, the groups had no significant
difference (p = 0.28).
Fig. 5 Prevalence of atypical Scheuermann's disease according to age.
Gender
When differentiating regarding gender, a significant difference was observed, with
a higher prevalence of atypical SD in male patients, representing 4.7% of the condition
among men and 1.1% among women (p < 0.01) ([Fig. 6 ]).
Fig. 6 Prevalence of atypical Scheuermann's disease according to gender.
Spine Level
When stratifying the levels of spinal involvement in atypical SD ([Fig. 7 ]), the T10-to-T12 level was the most prevalent one, representing 50% (n = 14) of
the total positive cases. For conducting the one-way ANOVA test and comparing multiple
groups with the linear contrasts test, the T9-to-T11 and the T8-to-T10 interval groups
were combined into a single group (T8–10/T9–T11). It was observed that there was a
significant difference in the total Cobb angle value between the 3 groups (p = 0.02, effect size = 0.68 [0.21–1.58]). Moreover, when the linear contrasts test
was performed, a significant difference was found between the T10-to-T12 and the T11-to-L1
groups (p = 0.046). In contrast, no significant difference was found when comparing other groups.
Fig. 7 Prevalence of atypical Scheuermann's disease according to the level of spinal involvement.
Scoliosis
When comparing the presence of scoliosis in patients with atypical SD according to
gender, there was no significant difference. Scoliosis among women was found in 14%
of the patients and 25% among the male patients (Fisher's exact test; p = 1) ([Fig. 8 ]).
Fig. 8 Comparison of scoliosis presence and patient sex in atypical Scheuermann's disease.
Schmorl Nodes
Of the 28 patients with atypical SD, 10 (37%) presented Schmorl's nodules on tomographic
examination ([Fig. 9 ]). No association was found between age groups or gender and the presence of Schmorl's
nodules (p = 1).
Fig. 9 Presence of Schmorl's nodes in patients with atypical Scheuermann's disease.
Discussion
In this epidemiological study of the Brazilian population, we found a prevalence of
atypical SD of 2.8%, utilizing the criteria defined by Sorensen.[2 ]
Since there is no uniform standard for the radiological diagnosis of the atypical
form of the disease, we chose to use the Sorensen criteria as it is the standard most
used in other previous research.[15 ]
[20 ]
The design of our study based on CT allowed us to identify atypical SD in patients
without clinical complaints, which would otherwise be missed in a study based on the
patient's clinical complaints. As the demand for the tomography exam was due to complaints
other than the spine, we reduced the presence of selection bias in our sample since
the selection of patients from a spine surgery office would include a selection criterion,
that is complaining of deformity or spine pain.
Typically, patients present Scheuermann's kyphosis during the bone growth phase of
adolescence.[5 ]
[7 ]
[8 ] Thus, all patients presenting the radiological characteristics of the disease already
had it at the time of our evaluation. Furthermore, to avoid the possibility of false
positive cases resulting from degenerative disc disease, we limited the age criteria
for including patients in our study. Therefore, none of the cases diagnosed with atypical
SD may correspond to patients with degenerative disc disease, and we have not failed
to identify any patient who has not yet developed the disease.
There is still controversy in the literature regarding the prevalence of SD according
to patients' gender. Thus, some studies showed a similar prevalence in both genders,[3 ]
[15 ]
[20 ]
[21 ]
[22 ]
[23 ] but most previous research demonstrated a higher incidence in men.[1 ]
[3 ]
[8 ]
[17 ]
However, none of these studies evaluated separately patients diagnosed with atypical
SD. Thus, despite agreeing with most studies, our results with a higher prevalence
of SD in males only represent patients with the atypical form of the disease.
Sorensen described radiographic criteria for typical SD, including anterior wedging
greater than 5° in at least 3 adjacent vertebral bodies.[2 ] Schmorl's nodes, irregular vertebral endplates, narrowed intervertebral disc, and
anteroposterior elongation of the apical vertebral bodies are other radiological features
associated with MS. On the other hand, Blumenthal et al.[24 ] described the criteria for atypical MS, including wedging in one or two vertebral
bodies, changes in the vertebral endplate, narrowing of the disc space, and anterior
Schmorl's nodes.
We chose to use Sorensen's criterion of anterior vertebral wedging of more than 5°
because it is a more objective criterion than simply using vertebral wedging as suggested
by Blumenthal et al.[24 ] We were also able to investigate the presence of scoliosis and Schmorl's nodules,
and just as previous studies have shown, the prevalence of scoliosis was around a
third of cases.[25 ] In around 40% of the exams, we could identify the presence of Schmorl's nodules,
which agree with the findings of Heithoff et al.,[12 ] according to which the presence of Schmorl's nodules would be just one of the diagnostic
criteria for atypical DS.
The main limitation of the present study lies in the fact that we used the CT exam
to identify the presence of SD since the exam of choice is the full spine X-ray. However,
as the diagnosis of atypical SD is based on the individual wedging of each vertebra,
the chosen exam did not affect our results. Furthermore, we reduced the possibility
of selection bias since patients underwent the examination for reasons other than
the spine.
Conclusion
Scheuermann's kyphosis is a frequent spinal deformity. Our study, obtained from a
large convenience sample of patients evaluated with abdomen tomography, allows a better
understanding of the actual prevalence of the atypical pattern of this condition.
The CT examination made it possible to study a population sample without selection
bias and to find a prevalence of 2.8% of atypical SD in the selected sample.
