Artikler fra Anvendt Sundhedsforskning
1: SEAS 2: Brace
1: n= 29, MA= 12.4 ±0.9
2: n= 24, MA= 12.3 ±0.8
1: 40 minuts supervised training x 1 per week and 15 minuts home-based exercises x 1 per day.
Furthermore, 1,5 hours of instructions in a personalized exercise program every 2-3 months
2: Gets instructed in wearing a brace 23 hours a day. The brace gets controlled and regulated after 1 month, and then every 3 months.
Follow-up: After 6 and 12 months
N/A Significant
reductions of Cobb angle in both groups after 12 months, however, brace seemed to have a superior effect.
1: No 2: N/A
Mean age (MA), Standard deviation (SD), and characteristics of the included studies are shown in the table.
*: Standard Error (SE), N/A: Not applicable
Bias
Monticone et al. 2014
Baseline
Mean ± SD 95%(CI)
P-value
baseline 42 months
Mean ±SD 95% (CI)
P-value
within P-value between
54 months
Mean ±SD 95% (CI)
P-value
between Low risk IG 19.3 ± 3.9 (18.25-20.35) 0.861 14.0 ± 2.4 (13.35-14.65) 14.3 ± 2.3 (13.68-14.92) <0.001#¥
CG 19.2 ± 2.5 (18.53-19.87) 20.9 ± 2.2 (20.31-21.49) 22.0 ± 1.6 (21.57-22.43)
Kuru et al.
2016
Baseline
Mean ± SD 95% (CI)
P-value
baseline 6 months
Mean ±SD★ 95% (CI)
P-value
within P-value between
High risk IG 33.4 ± 8.9 (28.47-38.33)
0.397
32.3 ± 7.2 (28.31-36.29)
0.003# CG1 30.3 ± 7.6 (26.09-34.51) 33 ± 6.9 (29.18-36.82)
CG2 30.3 ± 6.6 (26.64-33.96) 33.8 ± 7.2 (29.81-37.79)
Kim &
HwangBo 2016
Baseline
Mean ± SE 95% (CI)
P-value
baseline 12 weeks
Mean ±SE 95% (CI)
P-value
within P-value
between Moderate
risk IG 23.6 ± 1.5 (20.30-26.90) 12.0 ± 4.7 (1.66-22.34) <0.05# <0.05#
CG 24.0 ± 2.6 (18.28-29.72) 16.0 ± 6.9 (0.81-31.19) <0.05#
Ko & Kang 2017
Baseline
Mean ± SD 95% (CI)
P-value
baseline 12 weeks
Mean ±SD 95% (CI)
P-value
within P-value between
High risk
IG T:14.46 ± 2.07 T: 14.33 ± 2.06
<0.001#∆
L: 15.95 ± 1.84 L: 15.21 ± 1.91
CG T=13.93 ± 1.96 T: 13.82 ± 2.04
L=15.36 ± 2.07 L: 15.20 ± 1.98
Kumar et al. 2017
Baseline
Mean ± SD 95% (CI)
P-value
baseline 12 months
Mean ±SD 95% (CI)
P-value within
P-value
between Moderate
risk IG 12.61 ± 1.81 (11.71-13.51) 0.83 6.83 ± 1.72 (5.97-7.69) <0.001# <0.001#
CG 12.72 ± 1.40 (12.02-13.42) 9.67 ± 1.32 (9.01-10.33) <0.001#
Zheng et al.
2018
Baseline
Mean ± SD 95% (CI)
P-value
baseline 12 months
Mean ±SD 95% (CI)
P-value
within P-value
between High risk
IG 27.0 ± 3.6 (25.63-28.37) 0.334 24.79 ± 4.36 (23.13-26.45) 0.03# 0.039# CG 28.0 ± 3.6 (26.48-29.52) 22.13 ± 4.78 (20.11-24.15) <0.001#
Mean, standard deviation (SD), Standard Error (SE) and p-values are stated for Cobb angle at baseline and follow-up. Confidence intervals (CI) are calculated by mean ± t-value*SE
IG: Intervention group (the exercise method which is investigated), CG: Control group (every other intervention and observation compared), #: Significant result,
★: Values are stated in the difference from baseline, without standard deviations, : Values are stated with SE instead of SD, T: thoracic spine, L: lumbar spine, ¥: Not stated whether the p-value is after post-intervention or follow-up, ∆: P-value is only significant in the lumbar spine
Participant characteristics
The total amount of participants in all studies was 304, of which 144 participants was in the intervention groups. There was a larger proportion of girls in the six studies. The mean age varied from 11.6 to 15.6 years.
Likewise, the size of the accepted Cobb angles varied from 10 to 60 degrees, respectively (Table 1).
Content of interventions
The participants in three of six studies [20,22,23] received general exercises, as for instance active self-correcting-, task oriented- and core stabilization exercises. In the three remaining studies [19,21,24], the participants received PSSE such as “the Schroth method” and “SEAS” (Table 1).
Between studies, there were great differences in duration and intensity of the interventions. The intervention with the shortest duration was 12 weeks, while the interventions within the longest was averagely 42 months (until participants reached full skeletal maturity). A combination of supervised- and home-based exercise program was seen through the intervention period, with the exception of two studies, which exclusively used supervised exercises.
The interventions in the control groups were primarily physical exercise, however, the control group in the study by Zheng et al. [24] received brace as a treatment. It is uncertain which kind of treatment the control group in the study by Ko & Kang [22] received.
Effect of physical exercise on the Cobb angle
The results of the measurements on the Cobb angle at baseline and follow-up, confidence intervals (CI) at baseline and follow-up, and the quality assessment is illustrated in Table 2. CI are calculated for all studies by the authors of the current study, as these intervals were not displayed in any of the included studies. It was not possible to calculate CI in the study by Ko & Kang [22] due to missing information of these values in the article. Kuru et al. [21] did not provide mean and SD in their study, which is necessary to calculate CI. These values are found eligible in Burger et al. [10].
With respect to the Cobb angle, the participants within the studies were homogeneous at baseline.
There was a great variation of the degree of the Cobb angle throughout the studies at baseline, and also a great variation of the Cobb angle by the end of the interventions.
Despite this, all studies demonstrated a significant decrease of the Cobb angle after the intervention period.
Noteworthy, a decrease of Cobb angle was not solely seen in the intervention groups, as the control group in the study by Zheng et al. [24] displayed a larger decrease of Cobb angle than the intervention group.
Similarly, in the study by Kumar et al. [23], the control group which had received active self-correcting exercises, without any task-oriented aspects, gained an improvement of the Cobb angle. However, the improvement of the Cobb angle was significantly smaller than in the intervention group.
Quality assessment of studies
Most of the studies reported randomization and allocation. However, a risk of detection bias was seen in four of six studies [19,21,22,24]. The study by Zheng et al. [24] has a risk of attrition bias due to a high drop-out rate in the control group compared to the intervention group. Furthermore, it is unknown whether intention to treat (ITT) analysis are applied on these drop-outs. Similarly, it is unclear if there are drop-outs in the study by Ko & Kang [22], and how these drop-outs are handled. In addition, there is a risk of reporting bias in the studies by Ko & Kang [22] and Kuru et al. [21] (Figure 2).
Figure 2: Risk of Bias summary
Meta-analysis
The meta-analysis is a quantitative pooling of the results of the selected studies. Results from the studies of Zheng et al. [24] and Ko & Kang [22] are not included, as the interventions in the respective control groups were not comparable to the other studies. Kuru et al. [21] is the only study with two control groups, of which one of the groups are only being observed. This group is not included. Values for the study by Kuru et al. [21]
are found eligible in Burger et al. [10].
Figure 3: Forest plot of the results of the mean difference of the Cobb angle between control and intervention groups.
Overall, there was a significant reduction of the Cobb angle in intervention groups compared to control groups, with CI ranging from -7.19 to -1.09 (p=0.008) (Figure 3).
The result of the GRADE assessment, based on the studies included in the meta-analysis indicates that, physical exercise seems to have a positive effect on reducing the Cobb angle, however, the likelihood that the outcome is substantially different, is high, due to risk of bias and inconsistency [17].
Discussion
Overall, physical exercise has a significant improvement of the Cobb angle in adolescents with AIS. Likewise, the meta-analysis displays that there is a significant reduction of the Cobb angle in the intervention groups compared to the control groups. Using the GRADE tool, this pooling of the data estimates that physical exercise seems to have a positive effect on reducing the Cobb angle, however, the likelihood that the results can vary substantially is high [17].
The result of the current systematic review corresponds well to the results of other studies [7–9], as all studies have found a positive effect of physical exercise on reducing the Cobb angle. However, the results of the current study are based exclusively on RCT studies, in contrast to others [7–9].
The latest four systematic reviews regarding this topic, all report a positive effect of PSSE on reducing the Cobb angle [10–13]. However, the study by Burger et al. [10] is focusing only on the effect of “Schroth exercises” on the Cobb angle in contrast to the current systematic review, assessing multiple kinds of physical exercise. The last three of the aforementioned reviews [11–13] combines physical exercise with brace treatment but does not separate these in the conclusion in contrast to the current systematic review, who seeks to minimize all confounders.
The primary outcome in the current study was the Cobb angle, owning to the fact, that this measurement method of the spinal curvature, often is described as golden standard when diagnosing scoliosis [2].
Nonetheless, there are some concerns regarding this method, when deciding on the most tilted vertebrae, and therefore there is a risk of miscalculations up to five degrees [25].
Noteworthy, half of the studies in the current review are of high risk of bias, which affects the overall result;
primarily concerning how the studies handled missing data, and if there were risk of reporting bias.
In the two studies, that most likely did not perform ITT on dropouts [22,24], the results can seem more positive than they truly are [26], assuming that physical exercise has an effect on reducing the Cobb angle, even if this is not the case.
Concerns of reporting bias are present in the study by Ko & Kang [22], where the results are stratified in a thoracic and a lumbar Cobb angle, although this stratification is not further elaborated. The insignificant improvement of the Cobb angle at follow up could be due to this stratification, not seen in other studies.
Likewise, there is a risk of reporting bias in the study by Kuru et al. [21], in which the data on the Cobb angle are analysed with non-parametric tests. This could indicate that data was not normally distributed and given the lack of explicit description in the study, the transparency is reduced.
As previously mentioned, some studies [19,21,22,24] had a risk of detection bias, as lack of blinding unconsciously might have affected the measurement of the outcome.
The CI in the study by Kim & HwangBo [19] are wide, implying an imprecise estimate on the effect of physical exercise in this study. The effect on the Cobb angle could just as well have been 20 degrees, as approximately 1 degree.
Motor skills training (MST) are often used by physiotherapists and can be split into formal and functional training. Formal training aims to optimize movement patterns and muscle activity, while functional training implements specific tasks where movements are combined in a context of daily living [27].
All the general exercise interventions use some form of formal MST, furthermore, the studies involve the functional aspect in their interventions [20,23]. In the studies by Ko & Kang [22] and the control group in Kim
& HwangBo [19], the content of the intervention is unclear, but seems to include only the formal aspect. The importance of adding the functional aspect is emphasized, when working with MST [27]. The lack of functional aspect could have an impact on the modest progress seen in the study by Ko & Kang [22], and also the setback in the control group in the study by Kim & HwangBo [19].
PSSE strives to achieve a 3D correction of the spine and that these corrections must be related to daily living, why PSSE could be considered as MST. Several factors, such as the knowledge of basic motor control indicates that similarities can be found between the general physical exercise principles and PSSE, and that the general physical therapist possesses the competences necessary to treat adolescents with AIS.
The treatments in all the intervention groups are based on the individual scoliosis, however, it is unclear whether this is the case in the study by Ko & Kang [22].
Multiple studies suggest that muscle imbalance, among other factors, has a great impact on the progression of the scoliosis [28,29]. Based on these theories, it could be relevant to focus the treatment on the individual scoliosis, due to the fact that the conditions of the active and passive tissues surrounding the joints are related to the individual adolescent.
In the study by Ko & Kang [22] the smallest, yet significant, improvement on Cobb angle is seen. This finding is difficult to explain, but it could be due to the fact that it is unclear if the treatment was focusing on the individual adolescents’ scoliosis.
Despite a significant overall difference between the intervention groups and control groups, a large statistic variation is seen between studies, implying a substantial impact on the overall results of the meta-analysis.
These variations make it difficult to estimate the degree of the Cobb angle at baseline in order to see a positive effect of physical exercise.
The final GRADE score was downgraded due to risk of bias, with one study having high risk of bias, two studies have some concerns due to risk of bias, and only one study was low risk of bias. Furthermore, it was chosen to downgrade the final GRADE score due to inconsistency based on the percentage of variance attributable to study heterogeneity. The evidence was not downgraded due to indirectness as well as publication bias.
Since the CI in the meta-analysis does not cross the no-effect line at zero, it was chosen not to downgrade the final GRADE score due to imprecision.
The result of the meta-analysis is considered clinical relevant, as the ability to maintain or improve the degree of the scoliosis curve, can be seen as a criterion for success when handling this group of adolescents with AIS [4].
However, it should be emphasized, that two studies with high risk of bias are not included in the meta-analysis [22,24], and therefore not included in the GRADE assessment. The final results of the current review could have appeared differently if the two studies had been included. The results of these two studies must still be taken into consideration in the overall conclusion, but the impact of these studies should be interpreted with caution.
The GRADE score indicates that physical exercise seems to have a positive effect on reducing the Cobb angle, however, the likelihood that the outcome, when implementing these results, would vary substantially, is high, due to risk of bias and inconsistency [17]. Despite the various methodological quality of the studies, physical exercise is still considered relevant as a possible intervention when treating AIS, as the advantages outweigh the disadvantages.
The current systematic review contributes to the existing evidence, due to the exclusion criteria regarding brace wearing. It became clear through the search of literature, that studies which investigates the effect of physical exercise are often combined with “standard care” which, amongst other things, consists of simultaneous brace treatment.
Today, bracing is the most common choice of treatment for AIS. Therefore, the participants in the six included studies may not be entirely comparable to the background population. To minimize confounders, brace as a simultaneous treatment was excluded. Furthermore, this study contributes with new knowledge from pooling results of RCT studies, however, some of these studies are of low methodological quality.
Limitations
One of the limitations of the current systematic review is the low methodological quality of the included studies. Furthermore, a larger number of studies assessing the effect of physical exercise on the Cobb angle, could make it possible to make a more exact pooling of the results, and thereby minimize the risk of high inconsistency. High methodological quality studies are required for robust conclusions.
Conclusion
The current systematic review of the literature including quality assessment, indicates that physical exercise seems to have a positive effect on reducing the Cobb angle in adolescents with AIS. The GRADE score indicates that physical exercise may have a positive effect on reducing the Cobb angle. However, the likelihood that the effect on the Cobb angle, when implementing these results, can vary substantially is high, due to risk of bias and inconsistency, and therefore the results should be interpreted with caution.
Further RCT studies of high methodological quality are needed to give a more distinct indication of the degree of the Cobb angle at baseline, where physical exercise seems to have a superior effect. Furthermore, in relation to this research, it would be relevant to investigate the association between the degree of scoliosis and Health Related Quality of Life.
Ethical Approval
Ethical approval was not required due to the study design of this paper.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Conflict of interest
There seems to be no conflict of interest.
References
[1] Asher MA, Burton DC. Adolescent idiopathic scoliosis: natural history and long term treatment effects.
Scoliosis 2006;1:1–10. https://doi.org/10.1186/1748-7161-1-2.
[2] Scherl S. Adolescent idiopathic scoliosis: Clinical features, evaluation, and diagnosis 2018.
http://www.uptodatefree.com/topic/adolescent-idiopathic-scoliosis-clinical-features-evaluation-and-diagnosis (accessed October 23, 2019).
[3] Wong C. Mechanism of right thoracic adolescent idiopathic scoliosis at risk for progression; a unifying pathway of development by normal growth and imbalance. Scoliosis 2015;10:1–5.
https://doi.org/10.1186/s13013-015-0030-2.
[4] Negrini S, Donzelli S, Aulisa AG, Czaprowski D, Schreiber S, de Mauroy JC, et al. 2016 SOSORT guidelines: orthopaedic and rehabilitation treatment of idiopathic scoliosis during growth. Scoliosis Spinal Disord 2018;13:1–48. https://doi.org/10.1186/s13013-017-0145-8.
[5] Adobor RD, Joranger P, Steen H, Navrud S, Brox JI. A health economic evaluation of screening and treatment in patients with adolescent idiopathic scoliosis. Scoliosis 2014;9:21.
https://doi.org/10.1186/s13013-014-0021-8.
[6] Berdishevsky H, Lebel VA, Bettany-Saltikov J, Rigo M, Lebel A, Hennes A, et al. Physiotherapy scoliosis-specific exercises – a comprehensive review of seven major schools. Scoliosis Spinal Disord
2016;11:1–52. https://doi.org/10.1186/s13013-016-0076-9.
[7] Negrini S, Fusco C, Minozzi S, Atanasio S, Zaina F, Romano M. Exercises reduce the progression rate of adolescent idiopathic scoliosis: Results of a comprehensive systematic review of the literature. Disabil Rehabil 2008;2008:772–85.
[8] Anwer S, Alghadir A, Shaphe A, Anwar D. Effects of Exercise on Spinal Deformities and Quality of Life
in Patients with Adolescent Idiopathic Scolisis. Hindawi Publ Corp 2015;2015:1–16.
[9] Fusco C, Zaina F, Atanasio S, Romano M, Negrini A, Negrini S. Physical exercises in the treatment of adolescent idiopathic scoliosis: An updated systematic review. Physiother Theory Pract 2011;27:80–
114. https://doi.org/10.3109/09593985.2010.533342.
[10] Burger M, Coetzee W, du Plessis LZ, Geldenhuys L, Joubert F, Myburgh E, et al. The effectiveness of Schroth exercises in adolescents with idiopathic scoliosis: A systematic review and meta-analysis.
South Afr J Physiother 2019;75:1–9. https://doi.org/10.4102/sajp.v75i1.904.
[11] Thompson JY, Williamson EM, Williams MA, Heine PJ, Lamb SE. Effectiveness of scoliosis-specific exercises for adolescent idiopathic scoliosis compared with other non-surgical interventions: a systematic review and meta-analysis. Physiotherapy 2019:214–34.
[12] Day JM, Fletcher J, Coghlan M, Ravine T. Review of scoliosis-specific exercise methods used to correct adolescent idiopathic scoliosis. Arch Physiother 2019:1–11.
[13] Laita LC, Cubillo CT, Gómez TM, del Barrio SJ. Effects of corrective, therapeutic exercise techniques on adolescent idiopathic scoliosis. A systematic review. Arch Argent Pediatr 2018;116:e582–9.
[14] WHO. Adolescent health. World Health Organ n.d.
https://www.who.int/southeastasia/activities/adolescent-health (accessed October 28, 2019).
[15] Cochrane Library. Cochrane PICO search | Cochrane Library n.d.
https://www.cochranelibrary.com/about/pico-search (accessed March 21, 2020).
[16] The Cochrane Collaboration. RoB 2: A revised Cochrane risk-of-bias tool for randomized trials n.d.
file:///Users/bruger/Zotero/storage/72BXDHGX/rob-2-revised-cochrane-risk-bias-tool-randomized-trials.html (accessed December 1, 2019).
[17] Schünemann H, Brozek J, Guyatt G, Oxman A. GRADE handbook 2013.
https://gdt.gradepro.org/app/handbook/handbook.html#h.33qgws879zw (accessed December 12, 2019).
[18] Cochrane Community. RevMan 5 n.d. https://community.cochrane.org/help/tools-and-software/revman-5 (accessed December 10, 2019).
[19] Kim G, HwangBo P. Effects of Schroth and Pilates exercises on the Cobb angle and weight distribution of patients with scoliosis. J Phys Ther Sci 2016;28:1012–5.
[20] Monticone M, Ambrosini E, Cazzaniga D, Rocca B, Ferrante S. Active self-correction and task-oriented exercises reduce spinal deformity and improve quality of life in subjects with mild adolescent
idiopathic scoliosis. Results of a randomised controlled trial. Eur Spine J 2014;23:1204–14.
[21] Kuru T, Yeldan İ, Dereli EE, Özdinçler RA, Dikici F, Çolak İ. The efficacy of three-dimensional Schroth exercises in adolescent idiopathic scoliosis: a randomised controlled clinical trial. Clin Rehabil 2016;30:181–90.
[22] Ko K-J, Kang S-J. Effects of 12-week core stabilization exercise on the Cobb angle and lumbar muscle strength of adolescents with idiopathic scoliosis. J Exerc Rehabil 2017;13:244–9.
[23] Kumar A, Kumar S, Sharma V, Srivastava RN, Gupta A, Parihap A, et al. Efficacy of Task Oriented Exercise Program Based on Ergonomics on Cobb’s Angle and Pulmonary Function Improvement in Adolescent Idiopathic Scoliosis- A Randomized Control Trial. J Clin Diagn Res 2017;2017:1–4.
https://doi.org/10.7860/JCDR/2017/27497.10335.