Abstract
Objective. To examine the frequency of impaired spinal mobility in patients with chronic back pain of short duration and to compare it with the frequency of impaired spinal mobility in patients with axial spondyloarthritis (axSpA), possible SpA, and no SpA.
Methods. The SpondyloArthritis Caught Early (SPACE) cohort includes patients with chronic back pain (≥ 3 mos, ≤ 2 yrs, onset < 45 yrs). Spinal mobility was assessed with lateral spinal flexion, chest expansion, cervical rotation, occiput-to-wall distance, and lumbar flexion. Hip mobility was assessed with intermalleolar distance. Mobility measures were defined as impaired if below the 5th percentile reference curve from general population, adjusted for age and height when appropriate. Proportions of patients categorized with impaired mobility were examined with chi square.
Results. In total, 393 patients with chronic back pain were included: 142 axSpA, 140 possible SpA, and 111 no SpA. Impairment in ≥ 1 mobility measure was present in 66% of all patients. The most frequently impaired mobility measure was lateral spinal flexion (40%), followed by chest expansion (22%), cervical rotation (18%), intermalleolar distance (17%), lumbar flexion (15%), and occiput-to-wall distance (11%). No statistically significant differences in proportion of patients with impaired spinal mobility were found between patients with axSpA and the other subgroups in any of the tests.
Conclusion. Two out of 3 patients with chronic back pain of short duration had impaired spinal mobility compared to the general population. Impaired spinal mobility occurs as often in patients with early axSpA as in other forms of chronic back pain.
Back pain is one of the most common musculoskeletal disorders, affecting up to 80% of the population at some point in life1 and representing large individual and societal costs1,2. Most patients with back pain experience a natural recovery within some weeks. However, symptoms and functional limitations remain over time in about 10%1, and some of these patients have an inflammatory rheumatic disease such as axial spondyloarthritis (axSpA)3,4,5,6. AxSpA is characterized by inflammatory back pain and progressive restriction in spinal mobility7, with available treatments allowing us to improve health-related quality of life through control of symptoms and inflammation as the primary treatment goal8.
Early initiation of treatment is considered favorable in the disease course of patients with axSpA, and the response to tumor necrosis factor inhibitor (TNFi) therapy may be better when initiated early in the disease course9,10. Further, regular exercise is included in the management recommendation throughout the disease course8 and is shown to reduce disease activity and improve spinal mobility11 even in patients receiving stable TNFi therapy12. Because effective treatment is available, it is important to recognize patients with inflammatory rheumatic disease among the large group of patients with chronic back pain.
Reduced spinal mobility is regarded as an important clinical feature of axSpA, emphasized by its inclusion in the core set of clinical assessment defined by the Assessment of SpondyloArthritis international Society (ASAS)13. Both structural damage and inflammation may contribute to spinal mobility impairment14,15. In early disease, impairment in spinal mobility is shown to be more influenced by inflammation, whereas in later disease, structural damage is also important14. Lateral spinal flexion and the frequently used index of spinal mobility, the Bath Ankylosing Spondylitis Metrology Index (BASMI), have the best ability to discriminate between patients with and without structural damage in ankylosing spondylitis (AS)15. However, according to the current ASAS classification criteria, patients with axSpA can be classified before structural, radiographic changes have occurred, and the role of impaired spinal mobility as a disease-specific clinical feature in the early phase of the disease is not clear16. Age-adjusted normal values have recently been defined for the spinal mobility measures, which enables comparing spinal mobility of patients with the general population17. The objectives of our study were therefore to examine the frequency of impaired spinal mobility in patients with chronic back pain of short duration; and further, to compare the frequency of impaired spinal mobility between patients with axSpA and those with other forms of chronic back pain with similar symptom duration.
MATERIALS AND METHODS
Patients
Data from the SpondyloArthritis Caught Early (SPACE) cohort are used for this analysis. The SPACE cohort is a European ongoing observational inception cohort established in 200918 and includes patients with chronic back pain for at least 3 months, not exceeding 2 years of duration, with an onset before the age of 45 years. All patients included in SPACE between January 2009 and December 2014 from the 5 rheumatology out-patients’ clinics are included in the current analysis: Leiden University Medical Center, Amsterdam Medical Center and Groene Hart Ziekenhuis, the Netherlands; University of Padua, Italy; and Diakonhjemmet Hospital, Norway. In the Netherlands, the SPACE protocol was approved by a medical ethics committee at Leiden University Medical Center (P08.105), in Norway by the regional committee for medical and health research ethics in South East Norway (2010/426), and in Italy by the Azienda Ospedaliera di Padova (2438P). The study was performed in compliance with the Helsinki Agreement. All patients provided their written informed consent before participation.
Assessments
Data used in the current study refer to the baseline visit, in which all patients underwent a diagnostic assessment for potential axSpA and were classified according to the ASAS axSpA criteria16. Laboratory assessment consisted of HLA-B27 typing, erythrocyte sedimentation rate (ESR), and C-reactive protein (CRP). The presence or history of SpA features including inflammatory back pain, arthritis, enthesitis, uveitis, dactylitis, psoriasis, inflammatory bowel disease, response to nonsteroidal antiinflammatory drugs (NSAID), and family history of SpA were recorded18. Imaging was obtained with plain radiographs of the pelvis with anteroposterior view and magnetic resonance imaging (MRI) of the sacroiliac joints with a semicoronal plane. All imaging was scored independently by 2 experienced readers in a central reading. The pelvic radiographs were scored according to the modified New York criteria19 and the MRI were scored according to the ASAS definition of active sacroiliitis20.
For the purpose of our study, patients were classified as having axSpA if they fulfilled the ASAS criteria for imaging and/or clinical axSpA16. Patients who did not fulfill the ASAS criteria were classified as possible SpA if they (1) presented with sacroiliitis on imaging with no other SpA feature, (2) were HLA-B27–positive with 1 other SPA feature, (3) had ≥ 1 of the following SpA features: peripheral arthritis, uveitis, dactylitis, heel enthesitis, psoriasis, inflammatory bowel disease, elevated ESR/CRP, or (4) had these 3 SpA features: inflammatory back pain, positive family history, and good response to NSAID. Patients were classified as no SpA if they had low possibility of having axSpA (e.g., either being HLA-B27–positive with no other SpA feature or having ≤ 2 of the following SpA features: inflammatory back pain, positive family history, or good response to NSAID). A detailed description of the categorization of patients in the possible SpA and no SpA groups has been published previously21.
Information was collected about the patients’ age, sex, age at onset of back pain and duration of back pain, location of back pain, and current use of medication. Body height and weight were measured and body mass index (BMI) was calculated (kg/m2); ≥ 25.0 was categorized as overweight. In addition, intensity of back pain during the last week was assessed by the patients on an 11-point numeric rating scale, anchored by 0 “no pain” and 10 “unbearable pain.”
Five measurements of spinal mobility and 1 measure of hip mobility were collected following the recommendations from ASAS13. Spinal mobility was assessed with lateral spinal flexion, chest expansion, cervical rotation, lumbar flexion measured according to the 10-cm Schober test and occiput-to-wall. Hip mobility was assessed with intermalleolar distance. All spinal mobility measurements were recorded in cm with 1 decimal, except for cervical rotation, which was recorded in degrees. Intermalleolar distance was recorded in cm and rounded to integer unit. The better of 2 tries for each measurement was recorded. More details on the measurement are in Supplementary Table 1 (available with the online version of this article).
The composite index BASMI includes lateral spinal flexion, cervical rotation, lumbar flexion, intermalleolar distance, and tragus-to-wall distance22. To calculate BASMI, the values of tragus-to-wall were derived from the occiput-to-wall results by adding 8 cm18. By doing so, the value of zero in the occiput-to-wall corresponds to 8 in tragus-to-wall used in the calculation of BASMI linear, both equivalent to no increased kyphosis23,24. Further, measures of tragus-to-wall and occiput-to-wall are known to be comparable across the entire scale24. The formula for BASMI linear was used to compute the total score, ranging from 0–10, in which the highest score represents the most impairment23.
Reference data from the general population for lateral spinal flexion, chest expansion, cervical rotation, intermalleolar distance, lumbar flexion, and BASMI were obtained from the MOBILITY study17. The mobility measures, except BASMI, were defined as impaired if they fell below the reference values for the fifth percentile curve from the general population, adjusted for age for all measures, and for chest expansion and intermalleolar distance also for height17. The BASMI has inverse scoring; therefore, reference values above the 95th percentile curve were defined as impaired. For occiput-to-wall, percentile curves could not be derived, and a cutoff of > 0 was considered impaired. To analyze the sensitivity, we also performed the analyses with the 2.5th percentile curve as the cutoff for impairment.
Statistical analyses
Patient characteristics are presented as mean with SD for continuous variables and as frequency (percentage) for categorical variables. Mobility measures and the BASMI are presented with mean (SD) for continuous variables with normal distribution or median with interquartile range for continuous variables with skewed distribution. Proportions of patients categorized with impaired spinal mobility are presented as frequency (percentage). Overall group differences (definite, possible, and no SpA) were examined with chi-square test for categorical variables, 1-way ANOVA for continuous normally distributed variables, and by Kruskal-Wallis test for continuous variables with skewed distributions. If statistically significant differences were detected in the overall group analyses, appropriate posthoc analyses (chi-square test with Yates continuity correction, Fisher LSD test, or Mann-Whitney U test) were applied. Statistical analyses were performed using SPSS version 21.0 (SPSS Inc.) and the figures were made in GraphPad Prism version 7.0 (GraphPad Software).
RESULTS
In total, 395 patients with chronic back pain were eligible, but 2 patients had insufficient information, so 393 were included in our present study. Of the included 393 patients, 142 (36%) fulfilled the ASAS axSpA criteria, 140 (36%) were classified as possible SpA, and 111 (28%) as no SpA. In the axSpA group, 58 (41%) fulfilled the imaging arm (28 with radiographic sacroiliitis and 30 with active inflammation on MRI) and 84 (59%) the clinical arm only of the ASAS axSpA criteria. In the possible SpA group, 108 (77%) who had normal imaging were HLA-B27–negative and had ≥ 1 of the following SpA features: peripheral arthritis, uveitis, dactylitis, heel enthesitis, psoriasis, inflammatory bowel disease, or elevated ESR/CRP. There were 20 HLA-B27–positive patients (14%) with normal imaging and 1 SpA feature. Eight (6%) who had normal imaging were HLA-B27–negative, with the following 3 SpA features: inflammatory back pain, positive family history, and good response to NSAID. Four patients (3%) had positive imaging and no SpA features.
The patients’ characteristics are shown in Table 1. Patients classified with axSpA were more frequently male (p < 0.001) and HLA-B27–positive (p < 0.001). They reported less back pain (intensity; p < 0.001) but more commonly buttock pain (p = 0.03), and less frequently lumbar pain (p = 0.01). There were no differences in age at onset or duration of back pain between the subgroups.
Characteristics of the patients according to subgroups. Values are n (%) unless otherwise specified.
Mobility measures are presented in Table 2. Comparisons between the subgroups showed that the axSpA group compared to possible SpA and no SpA groups had statistically significant better intermalleolar distance (p = 0.01) and better cervical rotation than the possible SpA group (p = 0.01). There were no differences between the subgroups in the other mobility measures. In the BASMI, the axSpA group had a lower score (better mobility) compared to the possible SpA and no SpA groups (p = 0.01).
Spinal mobility measures according to subgroups. Values are mean (SD) except where indicated.
The proportions of patients categorized with impaired spinal mobility are shown in Table 3. In all patients with chronic back pain, impairment in at least 1 mobility measure was present in 66% among those with complete assessment. The most frequently impaired mobility measure was lateral spinal flexion (40% of the patients), followed by chest expansion (22%), cervical rotation (18%), intermalleolar distance (17%), lumbar flexion (15%), and occiput-to-wall distance (11%). Twenty-nine percent of the patients were categorized with impaired spinal mobility according to the composite score BASMI.
Proportion of patients categorized with impaired spinal mobility according to subgroups. Values are n (%).
There were no statistically significant differences in the proportions of patients categorized with impaired mobility according to the single instruments among the chronic back pain subgroups in any of the mobility measures (Table 3). However, for the BASMI, in accordance with the scores, a statistically significant lower proportion of patients with axSpA (21%) had impaired mobility compared to possible SpA (33%) and no SpA (33%; p = 0.03). There were no statistically significant differences in the proportion of patients with impairment according to sex in any of the measures (data not shown). The distributions of lateral spinal flexion and BASMI in the subgroups are shown in Figure 1, with percentile curves illustrating the age-specific spinal mobility cutoff as defined from the general population. Details of other mobility measures are outlined in Supplementary Figures 1–5 (available with the online version of this article).
Distribution of measurements for lateral spine flexion (A) and BASMI (B) in function of age and with percentile curves derived from measurements in the general population. Patients below the fifth (A) and above the 95th (B) percentiles are categorized as having impaired spinal mobility. BASMI: Bath Ankylosing Spondylitis Metrology Index; SpA: spondyloarthritis.
In the subgroup of patients with axSpA, impaired mobility in at least 1 measure was present in 58% among those with complete assessment of mobility measures (Table 3). Among those with at least 1 impaired measure, lateral spinal flexion was most frequently impaired in 47 out of 76 (62%). Among the remaining 29 patients (38%), chest expansion was most frequently impaired in 14 (48%), followed by cervical rotation in 9 (31%), occiput-to-wall distance in 8 (28%), intermalleolar distance in 7 (24%), and lumbar flexion in 7 (24%; data not shown).
The proportions categorized with impaired mobility were compared between patients fulfilling the ASAS criteria for axSpA according to radiographic sacroiliitis, active sacroiliitis on MRI, and clinical arm, and no statistically significant differences were seen in any of the mobility measures. Sensitivity analyses with the 2.5th percentile curve as cutoff showed similar results (data not shown).
DISCUSSION
In this study of patients with chronic back pain of short duration, we have shown that spinal mobility was impaired in 1 or more mobility measures in 66% of patients and that the most frequently impaired mobility measure was lateral spinal flexion. However, mobility measures recommended for axSpA are as frequently impaired in patients with early axSpA as in those with other causes of chronic back pain.
To our knowledge, this is the first study comparing spinal mobility in patients with chronic back pain with age-adjusted percentile curves from the general population. Large variations in spinal mobility are demonstrated in the general population17,25,26. Even so, in our current study, 2 out of 3 patients with chronic back pain had impaired mobility in 1 or more measures, defined as below the fifth percentile of the general population. In a previous study, a similar comparison was made for patients with established AS27, with an even stricter cutoff between normal and impaired mobility (2.5th percentile). This study showed that 79% of the patients with AS had impaired spinal mobility compared to the general population27, reflecting that restricted spinal mobility is more prevalent in patients with established AS.
In the current study, the most frequently impaired mobility measure in the axSpA group was lateral spinal flexion (35%), followed by chest expansion (20%), while the other mobility measures had lower proportions of impairment (12–15%). Lateral spinal flexion has also been reported to be the most frequently impaired mobility measure in established AS27, but unlike in our early cohort, the second most frequently impaired measures were found to be lumbar flexion and tragus-to-wall/occiput-to-wall distance27. This finding may indicate that there is a different pattern of mobility impairment in early disease compared to late disease. It has been suggested that screening for impairment in mobility in AS can be done by assessing lateral spinal flexion and lumbar flexion27. However, our results do not support screening with only these measures, because 28% (21 out of 76) of patients with early axSpA with normal lateral spinal flexion and normal lumbar flexion had impairment in another mobility measure.
Because of the natural disease course of progressive restriction in spinal mobility in patients with axSpA28, it could be expected that impaired spinal mobility would occur more often among patients with early axSpA than in patients with other causes of chronic back pain. However, no differences in proportions of impairment of spinal mobility were found between patients with axSpA and patients with chronic back pain from other causes in any of the measures. Therefore, our results indicate that mobility measures recommended for monitoring axSpA are of limited diagnostic value. Correlations between levels of back pain and reduced spinal mobility have previously been demonstrated in patients with chronic back pain29,30,31, and because patients with axSpA reported less back pain than the other subgroups in our study, this may at least partly explain the lack of between-group differences. Further, we found differences in the location of back pain; a larger proportion of patients with axSpA have buttock pain and a lower proportion have lumbar pain. It is plausible that the location of back pain could influence spinal mobility measures. Further research is needed to explore which factors are associated with reduced spinal mobility in early axSpA.
Even though the proportion of patients with impaired spinal mobility was similar in the subgroups, patients with axSpA showed better cervical rotation and intermalleolar distance than patients with other causes of chronic back pain. The intermalleolar distance is a measure of hip abduction and is therefore not included in the spinal mobility measures recommended by ASAS13. However, intermalleolar distance is included in the BASMI22 and is therefore often reported as an outcome measure in studies with axSpA. Spinal mobility is known to be influenced by age17,25,26, height17,26, sex25,26, and BMI26 in the general population as well as in patients with axSpA32,33,34,35. In our population, patients with axSpA were younger, taller, more often men, and had lower BMI than patients with other causes of chronic back pain, which may have influenced the results. However, the categorization of patients with impairment was adjusted for age (and height) and we did not find any differences between the sexes. On the other hand, we were unable to do subgroup analyses by BMI, because the extreme groups according to BMI were too small, which is a limitation of our study.
Strengths of our study are the use of wide inclusion criteria (chronic back pain ≥ 3 mos ≤ 2 yrs, onset < 45 yrs), the inclusion of patients from several countries in Europe, and having followed a thorough examination with assessment recommended by the ASAS. However, being a multicenter study, several assessors have collected data, which is a limitation.
Because the cohort consists of patients with short symptom duration, it is possible that some patients not fulfilling the ASAS axSpA criteria at baseline might develop axSpA later. In this and previous publications from this cohort, we therefore grouped patients not fulfilling the criteria but with a higher likelihood of developing axSpA based on baseline features as having possible SpA. Included patients may not fully represent the chronic back pain population of young age in a community setting, and findings are therefore most applicable to rheumatologist outpatients’ clinics.
To be specific enough to identify all patients with definitely impaired spinal mobility, and sensitive enough to identify those with potentially impaired spinal mobility, we chose to report the fifth percentile as a cutoff. To our knowledge, a unified definition of impaired mobility has not been established, but this cutoff is in line with articles assessing normative values in the general population17,25,26. Moreover, data using the 2.5th percentile were very similar.
Spinal mobility is impaired in 2 out of 3 patients with chronic back pain who are young and have short symptom duration, but the frequency of impairment is similar in patients with early axSpA and those with chronic back pain from other causes.
Although spinal mobility may play an important role in clinical decision making and treatment evaluation in patients with axSpA, our results indicate that mobility measures are of no diagnostic value.
ONLINE SUPPLEMENT
Supplementary material accompanies the online version of this article.
Acknowledgment
We thank all participating patients, rheumatologists, researchers, readers of the imaging, and research assistants.
Footnotes
The Diakonhjemmet Hospital cohort was supported by the Norwegian Rheumatology Association and by an unrestricted research grant from AbbVie (IMM-10-0042).
- Accepted for publication March 23, 2018.
