To evaluate if exercise programs in trials for patients with ankylosing spondylitis (AS) have the potential for effectiveness.
To evaluate if exercise programs in trials for patients with ankylosing spondylitis (AS) have the potential for effectiveness.
A systematic literature search was performed and randomized trials examining the effectiveness of exercise programs for AS patients were analyzed according to 3 elements: whether the exercise programs were designed according to the American College of Sports Medicine (ACSM) recommendations for developing cardiorespiratory fitness, muscular strength, and flexibility; whether physiologic responses were properly measured; and whether adherence to programs was monitored.
Twelve trials with a total of 826 AS patients were evaluated. Five trials included cardiorespiratory exercise as a part of the exercise programs. One of these met the ACSM recommendations for intensity, duration, frequency, and length of the exercise period. This trial showed the greatest within-group improvement in aerobic capacity (effect size [ES] 2.19). Five trials included muscular strength training, but none measured the physiologic responses nor met the recommendations for improving muscular strength. Eleven trials included flexibility training, but the programs were poorly described overall. Small improvements in spinal mobility (ES range 0.02–0.67) were reported in all trials. Finally, 4 trials reported on participants' adherence to the exercise programs, but only 1 provided sufficient information to evaluate the possible influence of the adherence.
The quality of interventions in exercise trials for patients with AS can be improved. Future trials should also focus on measuring and reporting physiologic responses and adherence to exercise interventions.
Ankylosing spondylitis (AS) is an inflammatory rheumatic disease affecting mainly the axial skeleton and sacroiliac joints, causing characteristic inflammatory back pain. Overall, the prevalence is between 0.1% and 1.4% (1). The disease starts relatively early in life, commonly between ages 20–30 years, and men are more often affected than women, with a ratio of ∼2:1 (1, 2). The main clinical features of AS are back pain and reduced mobility caused by inflammation in the axial skeleton (especially sacroiliitis). Peripheral joint involvement is reported in approximately one-third of the patients, most often hip, shoulder, and knee joints, and AS may also be associated with extraspinal manifestations, i.e., enthesitis, anterior uveitis, bowel, and heart (1). AS may result in varying degrees of structural and functional impairments and reduced general health (3, 4).
Current recommendations for management of AS encompass appropriate medication and exercises as the 2 cornerstones of treatment (5). Due to the typical disease characteristics, the main focus for exercise prescription has traditionally been improvement or maintenance of spinal mobility. However, recent research has revealed that AS also is associated with increased risk of cardiovascular morbidity (6). Therefore, it is important that exercise interventions for patients with AS are designed to target other aspects of physical fitness as well.
A Cochrane review including 11 randomized controlled trials (RCTs) concluded that exercises have small, but beneficial, effects on spinal mobility, physical function, and patient global assessment for patients with AS (7). However, the quality of the exercise programs, in terms of content and delivery, was not evaluated.
Physiologic responses to exercise are determined by the mode (frequency, intensity, and duration) of exercise (8). Therefore, exercise programs prescribed for disease management in AS should be in accordance with current evidence for developing aerobic capacity, muscular strength, and flexibility. Moreover, a dose-response association exists for exercise, just as it does for drugs, and the effect of exercise is therefore dependent on the patients' adherence to the prescribed programs.
It is reported that conclusions from earlier research can be used to assess the quality of the design of interventions, i.e., if current research provides clear indications of how specific interventions should be administered, such research can be used as a benchmark for assessing the quality of interventions (9). Therefore, the aims of this study were the following: 1) to evaluate whether the exercise programs in trials for patients with AS were in accordance with current well-recognized recommendations, 2) to explore whether physiologic responses were properly measured, and 3) whether adherence to exercise programs was monitored in the studies.
Studies were included if the study group was comprised of men and women with a diagnosis of AS according to the classification system described in the New York criteria (10). The study design should be RCTs, and the intervention should be any type of exercise program. Only publications available in the English language were included.
For the Cochrane review (7), relevant studies were identified by searching the Cochrane Central Register of Controlled Trials, Medline, EMBase, AMED, CINAHL, and PEDro up to January 2007. The search strategy recommended in the Cochrane Reviewers' Handbook 4.2.6 was applied (11). The reference lists of retrieved studies were scanned to identify additional relevant trials. When needed, authors were contacted for additional information.
For the purpose of this review, an updated search was performed (from January 2006 up to November 2009) in PubMed, EMBase, CINAHL, and PEDro.
The quality of the exercise programs was analyzed according to 3 elements. First, the content of the programs was evaluated, i.e., whether the program was theoretically founded to have the potential of gaining physiologic training effect. The different elements of the exercise programs were assessed according to the American College of Sports Medicine's (ACSM) recommendations for developing and maintaining cardiorespiratory fitness, muscular strength, and flexibility in healthy adults (12) (Table 1). Second, we evaluated whether physiologic responses were properly measured. Third, we evaluated whether adherence to the exercise program was monitored by use of patient-administered diaries or therapists' protocol. Within-group changes from baseline to after intervention were calculated and presented as effect sizes (ES), i.e., the change-score divided by the SD of the baseline score.
|Cardiorespiratory fitness||3–5||55–90% of maximal heart rate||20–60 minutes|
|Muscular strength||2–3||8–12 repetitions (10–15 for older/more frail persons)||Minimum: 1 set, 8–12 exercises|
|Load: last repetitions should be last ones performed before fatigue|
|Flexibility||2–3||10–30 seconds, 4 times/muscle group|
From the previously published Cochrane review of 11 studies, 10 were included in the present analyses, while 1 was excluded due to the type of intervention (not exercise) (13). From the updated literature search, 7 more recently published studies were considered. Five of these were excluded due to study design (not RCTs) (14–18). Finally, 12 RCTs of exercise programs for patients with AS were included in this review (14, 19–29).
Four of the studies compared different types of exercise programs with no intervention (19, 21, 24, 29). Four studies compared supervised exercise programs with unsupervised exercise programs (20, 23, 25, 26), and 4 studies compared 2 or 3 different exercise programs (14, 22, 27, 28). Table 2 shows an overview of elements of exercise programs (cardiorespiratory fitness, flexibility, and muscular strength), comparisons, and outcome measures for each study.
|Author, year (ref.)||Comparison†||Interventions||Outcome measures|
|Kraag et al, 1990 (19)||1||x||x||x||x|
|Hidding et al, 1993 (26)||2||x||x||x||x||x|
|Helliwell et al, 1996 (27)||3||x||x|
|Van Tubergen et al, 2001 (28)||3||x||x|
|Sweeney et al, 2002 (29)||1‡|
|Analay et al, 2003 (20)||2||x||x||x||x||x|
|Lim et al, 2005 (21)||1||x||x||x|
|Fernandez-de-Las-Penas et al, 2005 (22)||3||x||x||x|
|Altan et al, 2006 (23)||2||x||x|
|Ince et al, 2006 (24)||1||x||x||x||x|
|Cagliyan et al, 2007 (25)||2||x||x||x|
|Karapolat et al, 2009 (14)||3||x||x||x||x|
Cardiorespiratory training was a part of the exercise program in 5 trials (14, 20, 24, 26, 28). In addition, increased endurance was described as a treatment objective in 1 trial, but the exercise program was not described thoroughly (26). All the programs included activities involving large muscle groups, therefore meeting the recommendations according to type of activity.
In 2 of 5 trials, the participants' intensity was controlled during training (14, 24). The frequency and duration of the cardiorespiratory exercise was optimal in 4 trials (14, 20, 24, 28), but only 2 trials met the recommendations according to length of the exercise period (24, 26).
Conclusively, 5 of 12 trials included cardiorespiratory exercise as a part of the exercise programs, but only 1 of these trials met all of the recommendations for intensity, duration, frequency, and length of the exercise period (24) (Table 3).
|Author, year (ref.)||Activity, any using large muscle groups||Intensity, 55–90% maximum heart rate||Duration, 20–90 minutes†||Frequency, 3–5/week||Exercise period, 12–15 weeks||Physiologic responses, aerobic capacity|
|Hidding et al, 1993 (26)||Sports||Not reported||60||1||9 months||Supervised %: +4‡|
|Unsupervised %: ÷1|
|Van Tubergen et al, 2001 (28)||Walking||Not reported||30||5||3 weeks||Not reported|
|Analay et al, 2003 (20)||Bicycling||Low/not controlled||15–30||3||6 weeks||Supervised: ES 0.09|
|Unsupervised: ES −0.03|
|Ince et al, 2006 (24)||Aerobics||Low/controlled||30||3||12 weeks||Exercise group: ES 2.19|
|Control group: ES −0.35|
|Karapolat et al, 2009 (14)||Walking/swimming||Controlled, 60–70% of maximum heart rate||30||3||6 weeks||Walking: ES 0.65|
|Swimming: ES 0.51|
Cardiorespiratory fitness was assessed in 4 trials (14, 20, 24, 26), with a submaximal static bicycle test in 3 trials (20, 24, 26) as well as a maximal walking test on treadmill including direct measurement of oxygen uptake in 1 trial (14).Three trials reported a positive effect on cardiorespiratory fitness (14, 20, 24), with a range of changes (ES 0.09–2.19, baseline to after the intervention). Deterioration was reported for the unsupervised groups in 2 trials (20, 26). One trial met all the ACSM recommendations, and this trial reported the largest improvement in cardiorespiratory endurance (24) (Table 3).
Only 1 of 5 trials reported the participants' adherence to the exercise protocol thoroughly (by use of exercise diaries and by the group supervisor) (26). Hidding et al (26) reported that the mean attendance for group exercise was 62% and, on average, participants exercised at home for 1.8 hours per week. Without describing the method for registration of adherence to the exercise program, Ince et al reported that all of the participants regularly attended the exercise program (24).
Muscular strength training was a part of the exercise program in 5 of the included trials (20–22, 25, 26). The majority of these trials provided poor or no descriptions of the muscular strength exercises, and resistance was not dosed according to one repetition maximum in any of the 5 trials. Moreover, external load was not used in any of these trials, indicating low resistance. Further, only 1 trial reported the dose of muscular strength exercises in terms of repetitions and sets (22). Hence, it is difficult to evaluate whether the dose of muscular strengthening exercises was optimal.
The frequency of muscular strength training was optimal in 2 trials (20, 25), and 3 of 5 trials met the recommendations according to length of the exercise period (22, 25, 26). Conclusively, 5 of 12 trials included muscular strength training as a part of the exercise program, but the ACSM recommendations for developing muscular strength were not met in any of these trials (Table 4).
|Author, year (ref.)||Type of exercise||Load, 8–12 maximum repetitions†||Repetitions, 10–15||Sets, >1||Frequency, 2–3/week||Exercise period, 12–15 weeks||Physiologic responses|
|Hidding et al, 1993 (26)||Legs/trunk||NR||NR||NR||1||9 months||NR|
|Analay et al, 2003 (20)||Legs/arms/back||No external load||NR||NR||3||6 weeks||NR|
|Lim et al, 2005 (21)||NR||NR||NR||NR||7||8 weeks||NR|
|Fernandez-de-Las-Penas et al, 2005 (22)||3 exercises, abdominals, legs||No external load||8–10||2||1||16 weeks||NR|
|Cagliyan et al, 2007 (25)||NR||NR||NR||NR||2||12 weeks||NR|
Neither muscular strength nor any other relevant physiologic responses were measured in any of the 12 trials. Adherence to the muscular strength training was only reported in Hidding et al (details in the Cardiorespiratory Fitness section) (26).
Flexibility exercises, performed as dynamic exercises and/or stretching, were part of the exercise program and defined as an outcome measure in all trials except one (29). However, the descriptions of flexibility exercises were generally poor and were missing in 3 trials (19, 23, 28). These 3 trials were excluded from the results.
Seven trials described dynamic flexibility training as a part of the exercise program (14, 20–22, 25–27). Four of these included stretching exercises in addition to the dynamic exercises (14, 20, 22, 25), and 1 trial included only stretching exercises (24). The dynamic exercises and the stretching exercises are described separately below.
Back exercises were the primary focus in the 7 trials focusing on dynamic exercises, but some trials also included exercises for the whole body. Six of 7 trials met the recommendations according to frequency of the dynamic exercises (14, 20, 21, 25–27). The exercise period lasted for at least 12 weeks in 3 of 7 trials, therefore meeting the recommendations according to duration of the exercise period (22, 25, 26). The dose in terms of sets and repetitions of the dynamic exercises was only described in 2 trials (22, 27).
Stretching exercises were part of the exercise program in 5 trials (14, 20, 22, 24, 25). Two of these programs met the recommendations according to frequency and duration of the stretching exercise period (24, 25). Only 1 of 5 trials (22) described techniques, holding time, or number of repetitions, and the holding time of stretching in this trial exceeded the recommendations (Table 4).
In conclusion, 11 of 12 trials included flexibility training as part of the exercise program, and the majority of the trials met the ACSM recommendations according to frequency of the flexibility training. However, the descriptions of the exercises were insufficient or missing in all the included trials.
Flexibility/mobility was measured with relevant methods in 10 of 11 trials. All these trials reported small positive effects on flexibility after the exercise intervention (19–27), with changes ranging from ES 0.02–0.67 from baseline to after intervention (Table 5).
|Author, year (ref.)||Repetitions, >4||Stretch duration, 10–30 seconds||Frequency, 2–3 days/week (minimum)||Exercise period, 12–15 weeks||Physiologic responses, (spinal mobility)|
|Analay et al, 2003 (20)||NR||NR||3||6 weeks||Supervised: FFD ES 0.38|
|Schober test ES 0.31|
|Unsupervised: FFD ES 0.05|
|Schober test ES 0.24|
|Fernandez-de-Las-Penas et al, 2005 (22)||1–2||45 seconds to 4 minutes||1||16 weeks||Program 1: Schober test ES 0.15|
|Program 2: Schober test ES 0.46|
|Cagliyan et al, 2007 (25)||NR||NR||2||12 weeks||Supervised: FFD ES 0.67|
|Schober test ES 0.58|
|Unsupervised: FFD ES −0.02|
|Schober test ES −0.08|
|Ince et al, 2006 (24)||NR||NR||3||12 weeks||Exercise group: FFD ES 0.21|
|Schober test ES 0.11|
|Control group: FFD ES −0.04|
|Schober test ES −0.24|
|Karapolat et al, 2009 (14)||NR||NR||Once a day||6 days||Swimming: BASMI ES 0.27|
|Schober test ES −0.09|
|Walking: BASMI ES 0.14|
|Schober test ES −0.13|
|Conventional: BASMI ES 0.02|
|Schober: ES 0.0|
Three of 11 trials reported on adherence to the exercise protocol (24, 26, 28). Hidding et al (26) controlled and reported adherence (see details under Cardiorespiratory Fitness section). Van Tubergen et al (28) reported a mean attendance rate of 99% in the supervised groups, but adherence was not reported for the unsupervised group. Furthermore, as mentioned above, Ince et al (24) reported that all of the subjects regularly attended the exercise program without reporting any further details.
Evaluation of the quality of exercise programs in trials with AS patients indicates that the programs in general do not meet the recommendations for developing aerobic capacity, muscular strength, or flexibility. In one trial, the exercise program met the recommendations for duration, frequency, and length of the exercise period for cardiorespiratory exercise. This trial resulted in large improvement in aerobic capacity (24), while the remaining trials showed rather small change-scores after the exercise programs. In trials focusing on muscular strength, physiologic responses were not reported, while in flexibility trials the quality of interventions was poorly described. Future trials should also focus on measuring and reporting physiologic responses and adherence to exercise interventions for AS patients.
Physical activity and exercise are increasingly considered an important part of the treatment program for AS, together with appropriate medication (5, 30). Therefore, it is important that patients with AS are offered individually adapted exercise programs, and that the prescribed program is designed and delivered in accordance with current recommendations. A dose-response relationship exists for exercise as it does for medication. Prescription and monitoring of exercises as part of the disease management should therefore be as specific and accurate as for pharmacologic interventions. This review illustrates the importance of properly designed exercise programs, since programs that are in accordance with recommendations are more likely to result in beneficial effects for the participants.
Due to the typical clinical features of AS, the main focus for exercising has traditionally been improvement or maintenance of the spinal and thoracic flexibility. However, recent trials indicate that patients with AS also have increased risk of cardiovascular disease (CVD) (31–35); in prescription of exercise programs as part of disease management, this new insight should be taken into consideration. The ACSM and the American Heart Association recommend physical activity to reduce the risk of CVD (36). Regular participation in cardiorespiratory and muscular strength training is reported to be beneficial for prevention of CVD, partly by reducing the classic risk factors (37, 38). Further, cardiorespiratory exercise has been shown to also increase flexibility in AS patients (39). Therefore, an enhanced focus on cardiorespiratory elements may give effects that are more beneficial from the exercise programs.
Recent research has shown that implementation of positive findings from clinical trials is slow, and that a barrier for implementation may be clinicians' ability to carry out the treatment based on the information provided in published reports (40). The Consolidated Standards on Reporting of Clinical Trials recommend that reports from RCTs should include precise details of the intervention intended for each group, along with how and when interventions were actually administered. Many of the reviewed studies provided insufficient descriptions of the exercise intervention, therefore making it difficult to ensure that the exercise programs were designed according to recommendations, i.e., with sufficient dose (intensity and frequency) to achieve specific physiologic targets.
The way the intervention is assessed and reported is likely to influence the results of clinical trials (9). Patients' adherence with exercise programs can be controlled with the use of patient-administered diaries or therapists' protocol. Few of the studies in this systematic review reported on the participants' adherence to the exercise program, and only 1 of these studies provided sufficient information to evaluate the possible influence of the adherence. According to the results in the reviewed studies, it seems that supervised exercise programs are more likely to be effective than unsupervised individualized models (Tables 3 and 5). This may be explained by better adherence to supervised programs, but may also be due to better quality of supervised exercise programs in terms of control of intensity and duration. Exercise programs imply multiple demands on the participants, requiring attendance, motivation, and time expenditure, and the success is determined by the adherence to the prescribed program. Therefore, since the dose of exercise is decisive for the effect, the success or failure of an exercise program cannot be determined without information on the degree of adherence.
High-quality studies of the effects of exercise for patients with AS are warranted, but it is important that future studies provide exercise programs that are planned in accordance with recommendations for developing physical fitness. Further, relevant measures for evaluating physiologic responses should be applied, and it is essential that participants' adherence to the exercise program is monitored. In order to facilitate implementation of effective programs into clinical practice, the description of interventions must provide sufficient information for reproduction of the exercise programs.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be submitted for publication. Dr. Dagfinrud had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Dagfinrud, Halvorsen, Vøllestad, Niedermann, Kvien, Hagen.
Acquisition of data. Dagfinrud, Halvorsen, Hagen.
Analysis and interpretation of data. Dagfinrud, Halvorsen, Vøllestad, Niedermann, Kvien, Hagen.