Abstract
Objective Although exercise therapy is safe, effective, and recommended as a nonpharmacological treatment for axial spondyloarthritis (axSpA), there is a lack of guidelines regarding type and dosage. Insufficient knowledge about physical and physiological variables makes designing effective exercise programs challenging. Therefore, the goal of this study was to simultaneously assess trunk strength, spinal mobility, and the cardiorespiratory fitness of patients with axSpA.
Methods In a cross-sectional study, 58 patients with axSpA (mean age 40.8 yrs, 50% male, mean symptom duration 10.3 yrs) performed maximal cervical and trunk mobility and isometric strength tests in all planes (using David Back Concept devices) and a maximal cardiopulmonary bicycle exercise test (n = 25). Mobility and strength data were compared to healthy reference data. Cut-off values for clinical cardiopulmonary exercise testing interpretation were used to judge normality. Patients were compared based on radiographic involvement and symptom duration.
Results Both strength (P ≤ 0.02) and mobility (P ≤ 0.001) were significantly lower for the patients with axSpA compared to the reference. Strength deficits were comparable between the radiographic and nonradiographic groups (P > 0.05, except trunk extension [P = 0.03]), whereas mobility showed higher deficits in the radiographic group (cervical extension [P = 0.02] and rotation [P = 0.01], and trunk extension [P = 0.03] and rotation [P = 0.03]), regardless of symptom duration. Similarly, symptom duration positively affected oxygen pulse (P = 0.03), relative anaerobic threshold (P = 0.02), and aerobic capacity (P = 0.02).
Conclusion In patients with axSpA, strength is more affected than mobility when compared to healthy controls. Likewise, mainly the metabolic component of aerobic capacity is impaired, affecting cardiopulmonary fitness. These findings indicate that future personalized exercise programs in patients with axSpA should incorporate exercises for cardiopulmonary fitness next to strength and mobility training.
Spondyloarthritis (SpA) comprises a heterogeneous family of clinically and genetically closely related inflammatory rheumatic disorders. Depending on the location of the predominant clinical manifestation, the concept of SpA is split into axial and peripheral disease.1
Axial SpA (axSpA) is clinically characterized by sacroiliitis and/or spondylitis with buttock or back pain. In most patients, the loss of spine and/or chest mobility over time are the most prominent symptoms.2 A previous metaanalysis showed that patients with radiographic axSpA (r-axSpA) had more impaired spinal mobility compared to those with nonradiographic axSpA (nr-axSpA).3 However, to our knowledge, the effect of radiographic presentation on strength and cardiorespiratory endurance has never been investigated.
Additionally, because of the possible degenerative progress of axSpA, symptom duration might have a negative effect on mobility, strength, and/or cardiopulmonary health.
Objective measurements of cardiorespiratory health in SpA are limited in the literature, with only a few studies using standardized protocols such as the (modified) Bruce protocol, which is a staged uphill walking test on a treadmill estimating maximal oxygen uptake.4-7 Others have relied on endurance tests performed on a treadmill or bicycle ergometer while measuring cardiorespiratory variables,8-15 reporting reduced cardiorespiratory capacity among patients with SpA.8,9,15-18 Restrictive pulmonary function has also been noted in some studies.7,17
Likewise, only a few studies objectively measured muscle strength in SpA. Peripheral muscle strength was quantified using several different devices, such as handheld dynamometers,9,11,19-23 a leg press,24 an isometric knee extension device,25 or an isokinetic dynamometer.14,22,26 Trunk muscle strength is, however, even less often assessed in this population. To evaluate the latter, some used endurance tests,19 whereas others used a handheld dynamometer20-23,27 or isokinetic or isometric strength devices.28-30 These findings on strength show overall lower muscular strength and endurance for patients with SpA compared to healthy controls.9,11,14,19-26,28-30
Even fewer studies have combined measurements of strength with a cardiopulmonary exercise test (CPET). Measurements of lower limb muscle strength or hand grip strength were combined with a CPET, either on a treadmill or bike.9,11,14,24 Although the predominant clinical symptoms in axSpA manifest in the spine, none of these studies measured trunk muscle strength and objective spinal mobility in combination with a CPET.
Both restricted mobility and diminished cardiopulmonary fitness compromise the quality of life of patients with axSpA. Therefore, exercise is considered to be a cornerstone in the management of axSpA, with demonstrated benefits on disease outcomes independent of pharmacological treatment.2
Although exercise, following general recommendations for improving and maintaining physical fitness, is proven to be safe and effective, no specific guidelines are available regarding the required type and dosage of exercises. According to Ortolan et al, existing studies on the benefits of different exercise forms for patients with SpA are too varied and subject to bias, making it impossible to draw definitive conclusions. Moreover, it is not possible to blind participants in exercise studies, resulting in evidence rated as unclear.31 Additionally, the suggestion to follow public health recommendations for physical activity in SpA, which include improving cardiorespiratory fitness, muscle strength, flexibility, and neuromotor performance, is equally nonspecific for this patient group.
The insufficient knowledge about characteristic mobility, strength, and cardiorespiratory endurance in these patients makes it difficult to design effective exercise programs targeting possible deficits and leads to unclear recommendations. Further, there is a lack of objective measures in axSpA on trunk strength and cardiorespiratory health, especially with regard to the combination of both these variables and the unknown effect of symptom duration and radiographic involvement on these variables. Therefore, the aim of this study was to objectively measure trunk strength and spinal mobility in combination with an assessment of the cardiorespiratory fitness status in patients with axSpA at the same timepoint.
METHODS
Study population. Participants were consecutively recruited from the outpatient Rheumatology Department of the Ghent University Hospital and AZ Alma Hospital of Sijsele, both in Belgium. All participants fulfilled the 2009 Assessment of Spondyloarthritis international Society (ASAS) classification criteria for predominant axSpA and were between 18 and 60 years old. Exclusion criteria were pregnancy, acute flare, having cardiovascular disease or chronic obstructive pulmonary disease, trauma or surgery in the 6 months prior to baseline, or having other coexisting diseases interfering with the ability to perform the testing.
For the strength and mobility measures, the healthy control data as provided by David Health Solutions (age- and sex-matched Finnish population) were used as the reference population. For the physiological variables, the values were compared to reference data in datasets of healthy adults, and cut-off values for clinical CPET interpretation were used to judge normality.32-37
The study was approved by the ethics committee of the Ghent University Hospital (EC2018/1029 – B670201837207) and all patients signed an informed consent form prior to participation.
Procedure. The test procedure for this study consisted of several parts, including measurement of trunk muscle strength and mobility using David Back Concept (DBC) devices (David Health Solutions), and performance of a maximal CPET on a bike ergometer.
• Questionnaires and Bath Ankylosing Spondylitis Metrology Index. Three self-administered questionnaires were completed, namely the Bath Ankylosing Spondylitis Disease Activity Index (BASDAI), Bath Ankylosing Spondylitis Functional Index (BASFI), and a cardiovascular risk questionnaire that collected information on age, sex, height, body weight, symptom duration, smoking history, and medication use. Additionally, spinal and hip mobility were assessed using the Bath Ankylosing Spondylitis Metrology Index (BASMI; consisting of cervical rotation, lumbar flexion, lumbar lateral flexion, intermalleolar distance, and tragus to wall distance) and chest expansion was measured.
• Maximal mobility and isometric strength. To measure maximal mobility and isometric strength of the trunk and cervical region, without compensation strategies, DBCs (https://davidhealth.com/products) were used. Previous studies have reported the DBC devices to be valid and reliable.38-40
The general aim for the mobility measurements was moving as far as possible, without compensation strategies in the standardized position on the device, to achieve localized movements in the neck/trunk region to specifically measure the targeted area.
Maximal isometric flexion, extension, lateral flexion, and rotation strength measurements were performed on the trunk and cervical spine on the dedicated DBC devices. Participants were asked to inhale before the start of the test and to exhale during the performance. The target curve, as shown on the device, was to perceive a maximal contraction after 2 seconds and to hold this for 3 seconds. Positioning was at an angle of 30 degrees (for left and right lateral flexion, flexion for extension strength, and left and right rotation in the frontal, sagittal, and transversal planes) except for trunk flexion strength, which was measured in a neutral position (0 degrees). The strength measurement was performed twice; if there was a difference of > 10% in maximum strength between the trials, a third assessment was conducted. The average of the trials was calculated for further analysis.
• CPET. Exercise tolerance was measured using cycle ergospirometry, which included direct pulmonary measurements of oxygen consumption (VO2), carbon dioxide production (VCO2), and ventilation (VE), and was monitored by a 12-lead electrocardiogram (ECG) and blood pressure (BP) measurements.41 The test was preceded by a 30-second ECG at rest, and an alarm was set at 85% of the predicted maximal heart rate (HRmaxpred). The testing protocol included a warm-up period of 3 minutes of loose cycling at 0 watts and an increase of 50 watts every 3 minutes until exhaustion. Patients were asked to maintain a rate between 65 to 75 rotations per minute (rpm) and indicate the rate of perceived exhaustion on the Borg scale. The test was stopped if certain criteria were met, including exponentially exceeding 85% of HRmaxpred, indicating ≥ 16 on the Borg scale, not attaining ≥ 60 rpm, systolic BP exceeding 230 mmHg, patients indicating they felt unwell or dizzy, or patients reporting muscle cramps.
Based on the first ventilatory threshold (VT1), VO2, VCO2, and HR, and both the peak (peak) and predicted (pred) values, different derived variables were calculated to gain more insight on limiting factors of exercise tolerance. These variables were respiratory condition (VO2peak/VO2pred), oxygen pulse (VO2/HR), ventilatory efficiency (VE/VCO2), mechanical efficiency (VO2/watt), and the anaerobic threshold (VO2VT1/VO2predpeak).
• Data analysis. Trunk strength and mobility of patients with axSpA were compared with healthy age- and sex-matched Finnish adults, used as the reference population in the DBC software package. Mobility and strength deficits were calculated based on the formula: deficit = (reference value – patient value) × 100/reference value. Additionally, comparisons were made between r-axSpA (also known as ankylosing spondylitis) and nr-axSpA, and between short (≤ 2 yrs) and long (> 2 yrs) symptom duration. Aerobic capacity was compared to predicted values.32-35
Statistical analysis was performed with SPSS 28 (IBM Corp.). Histograms, box plots, Kolmogorov-Smirnov test, Shapiro-Wilk test, and QQ plots were used to indicate the distribution of the data and to investigate the outliers and extreme measurements.
Given the nonnormal distribution of the data, the nonparametric Wilcoxon signed-rank test was used to compare mobility and strength between the 2 matched groups.
Mann-Whitney U tests were used to compare the patients with r-axSpA vs nr-axSpA, as well as for patients with axSpA with short symptom duration vs those with long symptom duration for strength deficits, mobility deficits, and cardiorespiratory outcome variables.
A significance level of α < 0.05 was used throughout the data analysis.
RESULTS
Fifty-eight patients with axSpA (29 men and 29 women) were included, with a mean age of 40.8 (SD 11.9) years and mean BMI (calculated as weight in kilograms divided by height in meters squared) of 24.6 (SD 4.0); 19 (32.8%) had radiographic sacroiliitis fulfilling the modified New York criteria. The mean symptom duration was 10.3 (SD 10.9) years. Forty-five patients had a long duration (> 2 yrs) and 9 had a short duration. Four patients could not define the start of their symptoms (Table 1).
Patient demographics and baseline characteristics (N = 58).
The mean BASDAI score of the population was 3.7 (SD 1.9), BASFI score 2.9 (SD 2.0), BASMI score 2.7 (SD 1.5), and the chest expansion 5.1 (SD 2.0) cm (Table 1).
As for medication use, 72.4% of the patients used nonsteroidal antiinflammatory drugs, 15.5% conventional synthetic disease-modifying antirheumatic drugs (DMARDs), 5.3% corticosteroids, and 29.3% biologic DMARDs (Table 1).
Only patients included at the University Hospital of Ghent proceeded to the maximal exercise tolerance test, meaning that 25/58 patients (43%, 13 men and 12 women) performed this test. This subgroup of patients exhibited similar baseline characteristics to the overall group. Cycling test data from 2 male patients had to be excluded since their results indicated that they were unable to perform the test maximally and quit the test before reaching a respiratory exchange ratio ≥ 1.1 or without reaching a VO2max plateau.
Both mobility and strength analyses (Table 2 and Figure 1) show significantly lower values for the patients with axSpA compared to the healthy reference population for the cervical and trunk region in all directions. When comparing the relative deficits between mobility and strength measurements, the deficits for strength are much larger than those for mobility, indicating that strength is more affected compared to mobility.
Mobility and strength in the axSpA group compared to a reference population.
Mean relative mobility and strength deficits in the group with axSpA. Deficits were calculated based on the formula: deficit = (reference value – patient value) × 100/reference value.
Strength deficits were comparable between the radiographic and the nonradiographic group except for trunk extension deficit, which was higher in the nonradiographic group (P = 0.03). In contrast, the radiographic group showed significantly higher mobility deficits for cervical (P = 0.02) and trunk extension (P = 0.03) and cervical (P = 0.01) and trunk rotation (P = 0.03; Figure 2). No significant differences for mobility and strength deficits were found between those with symptoms for > 2 years (mean age 33.1 [SD 10.7] yrs) compared to those with shorter symptom duration (mean age 43.6 [SD 11.4] yrs).
Comparison of (A) mobility and (B) strength deficits between the radiographic and the nonradiographic group. * P = 0.02. ** P = 0.01. *** P = 0.03. SpA: spondyloarthritis.
Table 3 presents the mean cardiorespiratory variables for 23 patients with axSpA. Of these patients, 12 (52%) patients showed a limited aerobic capacity (VO2peak/VO2pred < 80%). Seven (30%) showed mechanical inefficiency with VO2/watt > 11, and 6 (26%) patients had low anaerobic threshold (VO2VT1/VO2predpeak < 40%) indicating deconditioning. Two (8%) patients had a VE/VCO2 > 32 and 8 (32%) showed an abnormal oxygen pulse (VO/HR < 80%).
Exercise physiology variables (N = 23).
No significant differences were found between the radiographic (n = 6) and nonradiographic (n = 19) groups for these cardiorespiratory variables. In contrast, symptom duration (n = 18, long duration; n = 6, short duration; n = 1, missing duration) positively influenced the oxygen pulse (P = 0.03), relative anaerobic threshold (P = 0.02), and aerobic capacity (P = 0.02; Table 4).
Exercise physiology differences between short and long symptom duration.
DISCUSSION
The results of this study show both mobility and strength to be significantly impaired in patients with axSpA compared to a healthy reference population, with strength deficits being larger than mobility deficits, independent from symptom duration or the presence of radiographic sacroiliitis. As expected, certain mobility variables (especially extension and rotation) were worse in patients with r-axSpA; however, no effect was seen on strength and cardiorespiratory endurance.
The physiological variables, measured during the cardiorespiratory test, suggest a general status of deconditioning with peripheral muscle function limitations having a significant effect on the exercise tolerance results in several patients with axSpA. In particular, patients with shorter symptom duration show worse VE/VCO2, oxygen pulse, and aerobic threshold.
These main findings imply that patients with axSpA should be screened not only on mobility, but also on strength and cardiorespiratory health. When decreases in ≥ 1 of these variables are identified, the rehabilitation should include a personalized and individual tailored exercise program to optimize these variables.
The BASMI is the standard instrument used in daily practice to evaluate mobility in patients with axSpA.2 In contrast, there is no such tool available to measure strength in patients with axSpA. Ample studies have evaluated mobility, but strength has rarely been measured objectively. In our study, we revealed that strength could potentially be a more important early and treatable aspect to measure and follow-up in all patients with axSpA as compared to mobility, which seems impaired mainly in r-axSpA. Moreover, with treatment strategies aimed at intervention in early stages of disease, the mere measurement of mobility impairment may be outdated. Our results emphasize that strength might be affected earlier in the disease course and, therefore, before mobility is impaired. Since strength measurements are not part of the standard clinical examination, in contrast with the mobility measurements documented with the BASMI, weakening of strength can easily be missed. Moreover, the BASMI, although a well-accepted tool, still remains operator dependent. In contrast, the DBC devices measure mobility in a standardized way, reducing intrareader and interreader variability.
A previous case series, using the DBC devices, evaluated the effects on range of motion and isometric muscular strength of the trunk muscles in patients with r-axSpA. However, only 3 patients were included. Since this was an exercise intervention study, the results show an increase both in mobility and strength compared with healthy controls.29 Wang et al measured isometric trunk muscle strength (n = 38) using a Kin Com device, which allows measurement in different postures. Their results are in agreement with our findings, demonstrating a decrease in strength for different postures.30 Even though they demonstrated a deterioration of strength as r-axSpA progressed, our data did not confirm these findings in axSpA. This difference in observation may simply be a result of a different patient group, as Wang et al only included patients with r-axSpA, whereas our study included patients with both r-axSpA and nr-axSpA, representing the full spectrum of the disease.
In the literature, 2 studies showed decreased isokinetic and isometric trunk flexion and extension strength in patients with r-axSpA compared to healthy controls.22,28 This is in accordance with our findings. In contrast to our results, in which flexion shows a greater deficit than extension (89.8% [SD 5.1] vs 50.4% [SD 17.7]), Durmus et al found extension to be more impaired (43-51% vs 28-29%).28 This difference may be a result of the different devices used, including different measuring positions. Another study measuring strength with a handheld dynamometer in axSpA did not find a significant difference in isometric torso extension strength.21,23 Since handheld dynamometer measurements might be dependent on the strength and skills of the investigator, caution is warranted when interpreting these results. In contrast to the DBC devices, where patients are fixed in a position to minimize compensational movements, handheld measurements need a trained investigator’s eye to detect all possible compensation strategies a patient can use.
The peripheral strength measurements of the lower limbs showed reduced strength and endurance.25,27 Despite the fact that Haglo et al did not compare their findings to healthy controls, they indicated that a training program had a positive effect.24 Isokinetic quadriceps strength showed no significant differences between a patient group with rheumatoid arthritis and a patient group with SpA, but was responsible for a large part of the variance in linear regression models when predicting functional activities such as flat floor walking and stair-climbing, possibly indicating weakness as well.14
In our study, limb strength was not measured. However, the CPET results indicate the metabolic variables as the main limiting factor for endurance. This implies that the peripheral muscles were the cause for cessation of the test. Indirectly this can be interpreted as weakness of the lower limb muscles when cycling.
Findings from previous studies using a bicycle ergometer for the maximal endurance test confirm the lower CPET variables for patients with SpA when compared to healthy controls. This indicates a reduced aerobic capacity and exercise tolerance.8,9,15-18 Similar to our findings, Carter et al and Özdemyr et al reported leg fatigue as the main reason for cessation of the test, indicating that peripheral muscle function determines the aerobic capacity.9,15
The reported restrictive pulmonary function as indicated by Seçkin et al and Fisher et al7,17 could not be confirmed in our study population. This might be accounted for by the fact that both those studies included patients with confirmed r-axSpA, who are likely more severely affected than the patients included in our study.
In this study, we used the reference data for strength and mobility as provided by the manufacturer (David Health Solutions). Since this was a Northern European database, we felt this was appropriate. For the physiological variables, we used the predicted and commonly used cut-off values. However, including a healthy matched control group, tested on the same devices, would be even more advisable.
Although this study is the first to our knowledge to combine both objective trunk strength and mobility measurements and a maximal exercise tolerance test in patients with axSpA, future studies should include strength and endurance measurements of the lower limbs in addition to objectively measured trunk strength and mobility and cardiopulmonary variables. By establishing individual combined baseline data and investigating the effect of symptom duration and radiographic presentation, we aimed to unravel the relationship between these measurements in order to develop a basis for personalized and patient-tailored exercise programs for patients with axSpA that address all of these different outcomes.
In future work, it would also be advisable to collect structured data on physical activity (PA), exercise habits, and sport activities. The amount of PA has been shown to correlate well with disease activity42,43 and disease-related factors act as a barrier to PA.44 Moreover, it has been shown that patients with r-axSpA lack adherence to physical exercise.45 Being less active might, next to the disease-related variables, contribute to a decreased physical fitness. Collecting structured data on disease activity (eg, Ankylosing Spondylitis Disease Activity Score [ASDAS] and C-reactive protein) could help to establish the relationship between disease-related factors and barriers to PA.
The results of the current study indicate that strength is more affected than mobility in both r-axSpA and nr-axSpA, although both are significantly worse compared to healthy controls. Likewise, mainly the metabolic component of aerobic capacity has proven to be impaired, thus affecting cardiopulmonary fitness. These observations provided by combined CPET, force, and mobility evaluation indicate that individualized and patient-tailored programs should be developed for patients with axSpA.
ACKNOWLEDGMENT
We would like to thank the physical therapists from the rehabilitation department of the AZ Alma Hospital for their help with collecting data on the DBC devices, and the Department of Physical Medicine and Rehabilitation of the Ghent University Hospital for the use of their accommodation and devices.
Footnotes
S. De Mits and T.M. Willems contributed equally to this work.
The authors declare no conflicts of interest relevant to this article.
- Accepted for publication February 27, 2024.
- Copyright © 2024 by the Journal of Rheumatology
