Scoliosis, lateral curvature of the spine beyond 10°, affects 2% to 3% of the population, or an estimated 6 to 9 million people in the United States. Because of preventive advances in tuberculosis and polio, currently more than 80% of cases are adolescent idiopathic scoliosis (AIS), but the severe and life-threatening cardiac and pulmonary complications are overwhelmingly manifested in the older population.1,2 Degenerative adult scoliosis (DS) results from structural deterioration, generally with onset after 40 years and may be related to osteoporosis.3 Most AIS cases develop in infancy or early childhood but are generally discovered at 8 to 15 years of age. In infancy, scoliosis is equally common in males and females,1,2,4 yet females are 8 times more likely to progress to scoliosis requiring treatment.1,2,4 Our series contains 7 people whose scoliosis was discovered in youth but remained untreated until well after their 21st birthdays. These are cases of aged-out AIS, without structural deformities, which we have examined separately and statistically looked at with the DS group because of their age and in spite of their earlier dates of diagnosis. The study was conducted without financial support and was registered with Clinical Trials.gov under Manhattan PMR number: NCT03110965.
TREATMENT OF SCOLIOSIS
Scoliosis can be painful and may affect gait, posture, and other areas of physical functioning, lowering self-esteem,5–9 negatively affecting body image in teenagers,9 and severely reducing respiratory function with age.10 Recent studies predict between 1% and 7% annual increase in untreated scoliotic curves.10–13 Standard care currently observes curves from 10° to 25°, recommends bracing from 25° to 45°, and surgery for curves above 45°.13
In one study, bracing significantly impeded progression of 72% of patients' scoliotic curves beyond 50° versus 48% of controls14 but did not actually improve AIS. Another shows differentially greater beneficial effect on lumbar versus thoracic curves.15 However, studies evaluating bracing and other conservative therapies are inconsistent, and they must be regarded as inconclusive.16–22 Several small studies are optimistic about yoga-like approaches.16,20,21 Surgery involves fusion, wiring, and/or rods bringing 44% to 56% improvement, which is reported to advance to 80% reduction of the curves if performed with pedicle screws.23–26 However, there is substantial comorbidity, including restriction of spinal mobility, hardware malfunctioning, extra strain on vertebrae above and below the fusion, and pseudoarthroses. One study documented more than 50% of revision surgery following Cotrel-Dubouset intervention27 but current estimates are less than 5%. Surgery is performed 38 000 times annually in the United States, its cost varying from $125 000 to $250 000.3 If the mean cost is $187 500, then the total annual cost for scoliosis surgery in the United States would be $7 125 000 000.1,28
We evaluated the efficacy of regular home practice of 1 or 2 yoga poses designed to strengthen the convex side of lumbar, thoracolumbar, and thoracic curves in AIS and DS. It was undertaken after observing that the side plank pose, done with the convex side inferior, had reversed simple and complex scoliosis in a number of AIS and DS patients.29 Although these 2 types of scoliosis are quite distinct, this method appeared to benefit people in both categories.
METHODS AND MATERIALS
We examined 97 consecutive patients, 54 DS and 43 AIS, in a retrospective study at our private physical medicine and rehabilitation practice in New York City, located in a neighborhood of affluent and educated people. The study was approved by the Chesapeake institutional review board, November 22, 2013, Pro 00009038. We studied those who submitted initial and terminal scoliosis radiographs and who self-reported that they had performed the pose(s) at least 3.5 days per week for the entire study period. The Yoga Journal recommends 2 to 3 sessions per week, as does the American Academy of Physician Assistants. 30,31 Approximately 60% of people practicing yoga at home actually practice 3 to 4 times per week.32 Recent strength studies suggest that an alternate-day regimen for body power is highly effective,33 yet no study we found recommended less weekly sessions than that. On this basis, we felt that every-other-day practice was likely to represent the benefits of the yoga poses. Among our candidates, patients with previous spinal surgery, pregnancy, or concurrent musculoskeletal or neuromuscular or psychiatric disorders were excluded from the study.
A slightly modified classical Iyengar side plank pose was used29,33 for all DS and AIS patients with lumbar and thoracolumbar curves. Patients were additionally instructed to elevate their ribs, which is not part of the classical Iyengar technique34 (see Figure 1). The pose was also modified for other medical conditions such as weakness, severe arthritis, or shoulder instability (see Figure 2). Complex or “S-shaped” curves' upper components, thoracic and cervicothoracic curves, were treated by adding a second contralateral strengthening pose, the half-moon pose, that consisted of looping a belt around the free leg's foot, holding it with the free arm, which was positioned perpendicular to the torso, and bulging that part of the spine, generally the cervicothoracic spine, upward (see Figure 3).
All patients were referred for scoliosis radiographs before treatment. Their radiologists or orthopedic surgeons were asked to measure Cobb angles and send radiographs and measurements to our clinic. Patients were then taught the side plank pose and in the case of “S” or “inverted S” curves, the half-moon pose as well. Patients performed them for 20 to 30 seconds daily for 1 week, as a training period, and then performed them at least once daily, holding them each time for as long as possible. Then patients were asked to return to the same radiologists or orthopedic surgeons for a second set of scoliosis radiographs 6 months later. The radiologists or orthopedic surgeons read the Cobb angles and sent their measurements to us.
The authors remeasured the Cobb angles of all radiographs we were able to obtain and agreed to consult the original radiologist or orthopedic surgeon if our measurements differed from theirs by more than 5°.
Paired sample t tests were used to evaluate the significance of changes in the mean Cobb angle of both lower and upper spinal curves for the degenerative and idiopathic scoliosis groups.
The current study included 74 compliant patients out of 97, 54 DS and 43 AIS between the ages of 85 (DS) and 8 (AIS) years, all with lumbar or thoracolumbar curves. Three DS patients did not do the pose at least every other day, 1 had surgery in the interim, and 1 was unreachable at and after the terminal date. Eleven AIS patients did not do the pose, 6 did not have subsequent radiography, and the parent of 1 patient refused to give us the results of the subsequent radiography. Of the 74 patients we studied, data from 49 with DS and 25 with AIS were analyzed.
Mean DS age was 62.1 years (SD = 16.1); mean AIS age was 14.2 years (SD = 2.75). A total of 23 DS patients and 12 AIS patients had “S” curves. Degenerative scoliosis and AIS groups had 11 and 7 dextrolumbar curves while 25 DS patients and 12 AIS patients had dextrothoracic curves, respectively. For all patients, Cobb angle measurements made by the study and those made by the patient's radiologist or orthopedist were within 3°. At the time of their second scoliosis radiographs, compliant DS and AIS patients self-reported practicing the side plank pose nearly daily (mean: 5.9 and 6.1 days per week, respectively) for an average of 1.10 and 1.15 minutes, respectively (range: 30 seconds to 3 minutes). Two DS and 4 AIS patients did not improve. Two DS patients' curves actually increased during the study period.
As the Table shows, DS patients' lower (lumbar or thoracolumbar) curves decreased from a mean of 38.9° (SD = 23.1) at baseline to 29.7° (SD = 19.1) or 9.2° after a mean of 10.5 months, an improvement of 23.7% (P < .001), or 2.4% per month. Thoracic and cervicothoracic curves in 23 DS patients had a mean Cobb angle of 37.7° (SD = 22.3) at baseline, and 27.3° (SD = 18.3) at the follow-up assessment, declining 10.4° or 27.6% over a mean of 11.1 months (P < .001). These patients improved a mean 2.5% per month over that time period. See the Table and Figures 4 and 5.
Adolescent idiopathic scoliosis
The AIS group showed significant improvement in lumbar and thoracolumbar curve angles from baseline to the postyoga follow-up measurement. Adolescent idiopathic scoliosis patients' mean curves decreased from 31.9° (SD = 17.4°) to 21.0° (SD = 21.7), an improvement of 10.9°, or 34.2% (P < .001) over a mean of 9.4 months (SD = 2.83), a mean improvement of 3.6% per month. The mean baseline Cobb angle for the 12 AIS patients with upper curves was 31.9° (SD = 14.9) while the comparable angle after the yoga intervention was 25.4° (SD = 25.7), a reduction of 6.5°, or 20.4%; (P = .004). Mean time between these assessments was 5.8 months, yielding a mean of 3.5% per month improvement. See the Table and Figures 4 and 5.
In the course of this work, we encountered 2 patients with osteogenesis imperfecta type I. One was a 14-year-old schoolboy with levothoracolumbar curve of 75° and dextrothoracic curve of 90°, and 2-in leg length discrepancy with rod from a previously fractured, and now shorter left femur; the second was a 53-year-old female professor with a levolumbar curve of 74°, a thoracic curve of 88°, and 20% left lateral listhesis at L3-L4. Both were denied surgery because of their bones' inability to retain hardware, and both resolved to use the methods of our study. The boy's thoracolumbar curve showed improvement to 54° after 4 months (27%), but his thoracic curve did not improve at first. His thoracolumbar and thoracic curves were subsequently measured at 53° and 77° after 2 years, improvements of 21° and 13°, respectively, or 28.4% and 14.4%, respectively. The professor improved only to 66° in the thoracolumbar and insignificantly in the thoracic curve in the first year. In November of 2015, she began doing the pose 5 times daily. Seven months later, her curves went to 49° and 57°, a 33.7% and 35.2% improvement over the full 25-month period, respectively.
Exception that illustrates the rule
One 8-year-old girl with a 30° curve mistakenly did the pose with the concave side of her “C” curve downward. When she returned 4 months later, her curve had increased to 40°. We discovered the error and explained to her and her father that the convex side had to be downward. She returned again in another 4 months: her curve was then measured at 30°. In subsequent radiographs her curve has decreased to 20°.
In this study, we found significant improvements in the Cobb angle of the lower (lumbar or thoracolumbar) scoliotic curve among 49 consecutive compliant DS patients and 25 consecutive compliant AIS patients who had been prescribed a single yoga pose in the case of “C” curves and that pose as well as the half-moon pose in the case of “S” curves. The limited number of patients with secondary curves showed some benefit to the secondary curve as well, although it was not as dramatic. Interestingly, this occurred over a relatively short period of time. Among our 74 compliant patients, 10 DS and 9 AIS patients' lower curves declined from above to below the threshold for bracing (25°), and 1 and 2 had upper curves that declined similarly. Six and 2 patients had lower curves crossing down below the surgical limit, and 7 and 0 patients had upper curves making that transition for the DS and AIS groups, respectively. One AIS patient who had been a surgical candidate straightened his curve enough to no longer be a candidate for bracing.
Untreated scoliosis is believed to progress to more severe spinal curvature over time.10–13 We found 2.2% and 3.6% monthly improvements in the lumbar and thoracolumbar curves of DS and AIS patients and 2.5% and 3.5% monthly improvements in thoracic curves, respectively. Even in a study without controls, this suggests that the simple isometric strengthening of the convex sides of curves is effective and might eliminate the need for surgery or bracing in some patients.
A dramatic finding is that AIS patients' mean lumbar curves improved more than 50% as quickly as the DS patients' mean lumbar curves. Although firm conclusions cannot be drawn from unequal measurement periods, it is interesting that while surgery is contraindicated because of a growing spine's instability, the current method might actually exploit the instability of growing children's spines to correct their curves more efficiently if treated earlier. A recent study found bracing significantly more effective in the lumbar than the thoracic curves in AIS.15 This matter should be the subject of further study.
In addition, there may be added value for adolescents because the daily home practice of these poses is unlikely to raise the same psychological and self-esteem issues that occur with bracing as a treatment. This drop in self-esteem has not to our knowledge been studied in adults but is not unlikely to occur there as well. Furthermore, the self-imposed therapeutic maneuvers require no elaborate equipment, take less than 3 minutes daily, and are virtually without cost. Yoga involves no encumbrance or restriction of movement in daily life and no visible markers of practice. While the best surgical studies show mean 80% improvement for patients, there are still unmistakable subsequent restrictions of movement that follow. Our study could not determine how much total improvement would be seen if the side plank pose were carried out for a longer period of time, but noted no restrictions of any kind.
The relative ease and low financial and time expenditure might prompt some parents and children to begin this innocuous treatment earlier, for example, in patients with strong family histories and, for example, 8° curves. As a result, some scoliotic curves might never advance to the degree that required surgical correction or other types of treatment later in life. In addition, use of the side plank pose has no notable side effects apart from occasional and mild wrist and shoulder discomfort. We, therefore, believe that future studies of this intervention are warranted to determine the factors that promote success of the procedure, its end-points, and optimal duration.
A simplified analysis of how humans stand erect conceptualizes it as a tensegrity structure involving the symmetrical downward pull of the dorsal, abdominal, intercostal, and paraspinal muscles. Some scoliosis, then, including degenerative cases, could be explained by asymmetry in the forces these muscles exert on the spine. The spine will bend toward the stronger side, and thus, the muscles of the convex side may be weaker than their smaller-appearing, more contracted counterparts on the concave side. We speculated that the side plank pose might be useful for strengthening the convex side's quadratus lumborum, iliopsoas, transversus abdominus, oblique, intercostal, and paraspinal musculature, which, in turn, might straighten the spine.
To test this out, we did electromyography of the downward side and upward side of an adult DS patient with a mild S-curve while she was in each of the 2 poses. We used a Cadwell Sierra Wave II with settings at 100 μV per division, and 10 millisecond divisions, filters at 10 and 3000 Hz and 37-mm monopolar needles.
In the side plank pose with convex left lumbar side down, the left quadratus lumborum and multifidi were more active than the right, and both were still more active when the pose was done on the left forearm (see Figures 6 and 7). The same asymmetry held but to a lesser extent with thoracic paraspinals at T8.
In the half-moon pose with right convex thoracic side down, the right-sided teres minor, the serratus anterior and paraspinals in the upper thoracic spine were more active than the left (see Figure 8). This pose is done with left arm perpendicular to the torso pulling a belt attached to the left foot (see Figure 3). This asymmetry was present down to the T8 level, but at L2-L3, the quadratus lumborum and multifidi became more active in the left (upper) side.
These electromyographic results indicate asymmetrically greater exertion by the lumbar muscles of the convex (downward) side in the side plank pose, and the convex thoracic muscles (downward side) in the half-moon pose (Figures 6–9).
Drawing firm conclusions from small series is challenging. In this unfunded study, we lacked a control group and had only patient reports on adherence to the treatment. However, it is notable that the reductions in the Cobb angle observed in these DS and AIS patients are superior to those from all the therapeutic studies of conservative treatments we identified; the bracing studies often focus on stability, not improvement.14 An adequately powered randomized controlled trial is planned to address this limitation.
Future studies of yoga as a treatment of scoliosis would benefit from control groups, better compliance, and inclusion of Lenke classification,13 used to determine surgical suitability by measuring the primary (largest) curve. Use of this measure will help make these studies comparable to surgical studies. In addition, studies with longer follow-up periods are needed to clarify the relationship between duration of treatment to the degree of improvement, the possible side effects of prolonged treatment, and the points at which one might reduce or stop the treatment and yet retain the benefits. Because past investigations found that both bracing and surgery impact quality of life,5–9 including quality-of-life measures would further facilitate comparability to more traditional studies. This might help patients and, where applicable, their parents in making decisions about the most appropriate treatments of scoliosis.
Asymmetrically strengthening the convex side(s) of the lumbar and/or thoracic curve(s) by daily practicing the side plank and/or the half-moon poses with the alterations practiced in this study may significantly reduce scoliosis in AIS and DS, respectively. Because this method appears effective, innocuous, and nearly without cost, it may be beneficially applied to quite small curves and serves as initial treatment for larger curves. In this way, it may help reduce the personal disadvantages and risks of bracing and surgery for some patients. The availability of this method might help patients and, where applicable, their parents in making decisions about the most appropriate first treatment of scoliosis.
1. National Scoliosis Foundation. Information and support. http://http://www.scoliosis.org
/info.php. Accessed July 3, 2014
2. Linker B. A dangerous curve: the role of history in America's scoliosis screening programs. Am J Public Health. 2012;102(4):606–616.
3. University of Maryland Medical Center. Degenerative adult scoliosis. http://umm.edu/programs/spine/health/guides/degenerative-adult-scoliosis#ixzz2nmnJiava. Accessed July 3, 2014.
4. Stolinski L, Kotwicki T. Trunk asymmetry in one thousand school children aged 7-10 years. Stud Health Technol Inform. 2012;176:259–263.
5. Kinel E, Kotwicki T, Podolska A, Białek M, Stryła W. Quality of life and stress level in adolescents with idiopathic scoliosis subjected to conservative treatment. Stud Health Technol Inform. 2012;176:419–422.
6. Misterska E, Glowacki M, Latuszewska J. Female patients' and parents' assessment of deformity- and brace-related stress in the conservative treatment of adolescent idiopathic scoliosis. Spine (Phila Pa 1976). 2012;37(14):1218–1223.
7. Negrini S, Donzelli S, Dulio M, Zaina F. Is the SRS-22 able to detect Quality of Life (QoL) changes during conservative treatments? Stud Health Technol Inform. 2012;176:433–436.
8. Parsch D, Gärtner V, Brocai DR, Carstens C, Schmitt H. Sports activity of patients with idiopathic scoliosis at long-term follow-up. Clin J Sport Med. 2002;12(2):95–98.
9. Pinquart M. Body image of children and adolescents with chronic illness: a meta-analytic comparison with healthy peers. Body Image. 2013;10(2):141–148.
10. Fu KM, Smith JS, Polly DW Jr, et al Morbidity and mortality in the surgical treatment of 10,329 adults with degenerative lumbar stenosis. J Neurosurg Spine. 2010;12(5):443–446.
11. Chuah SL, Kareem BA, Selvakumar K, Oh KS, Borhan Tan A, Harwant S. The natural history of scoliosis: curve progression of untreated curves of different aetiology, with early (mean 2 year) follow up in surgically treated curves. Med J Malaysia. 2001;56(suppl C):37–40.
12. Weinstein SI, Dolan LA, Spratt KF, Peterson KK, Spoonamore MJ, Ponseti IV. Health and function of patients with untreated scoliosis: a 50-year natural history study. JAMA. 2003;289(5):559–567.
13. Lenke LG. Commentary: continuing the quest for identifying specific criteria for the progression of adolescent idiopathic scoliosis. Spine J. 2012;12(11):996–997.
14. Weinstein SL, Dolan LA, Wright JG, Dobbs MB. Effects of bracing in adolescents with idiopathic scoliosis [published online ahead of print September 19, 2013]. N Engl J Med. 2013;369(16):1512–1521. doi:10.1056/NEJM oa1307337.
15. Thompson RM, Hubbard EW, Jo C-H, Virostek D, Karol LA. Brace success is related to curve type in patients with adolescent idiopathic scoliosis. J Bone Joint Surg Am. 2017;99:923–928.
16. Romano M, Minozzi S, Bettany-Saltikov J, et al Exercises for adolescent idiopathic scoliosis. Cochrane Database Syst Rev. 2012;8:CD007837.
17. Weiss HR, Werkmann M. Rate of surgery in a sample of patients fulfilling the SRS inclusion criteria treated with a Chêneau brace of actual standard. Stud Health Technol Inform. 2012;176:407–410.
18. Szwed A, Kołban M. Results of SpineCor dynamic bracing for idiopathic scoliosis. Stud Health Technol Inform. 2012;176:379–382.
19. Aulisa AG, Guzzanti V, Perisano C, Marzetti E, Falciglia F, Aulisa L. Treatment of lumbar curves in scoliotic adolescent females with progressive action short brace: a case series based on the Scoliosis Research Society Committee Criteria. Spine (Phila Pa 1976). 2012;37(13):E786–E791.
20. Bettany-Saltikov J, Parent E, Romano M, Villagrasa M. Physiotherapeutic scoliosis-specific exercises for adolescents with idiopathic scoliosis. Eur J Phys Rehabil Med. 2014;50(1):111–121.
21. Donzelli S, Lusini M, Zaina F. Characteristics of patients with more than 20° of improvement or worsening during conservative treatment of adolescent idiopathic scoliosis. Stud Health Technol Inform. 2012;176:354–717.
22. Miller DJ, Franzone JM, Matsumoto H, et al Electronic monitoring improves brace-wearing compliance in patients with adolescent idiopathic scoliosis: a randomized clinical trial. Spine (Phila Pa 1976). 2012;37(9):717–721.
23. Basu S, Rathinavelu S, Baid P. Posterior scoliosis correction for adolescent idiopathic scoliosis using side-opening pedicle screw-rod system utilizing the axial translation technique. Indian J Orthop. 2010;44(1):42–49.
24. Kelly DM, McCarthy RE, McCullough FL, Kelly HR. Long-term outcomes of anterior spinal fusion with instrumentation for thoracolumbar and lumbar curves in adolescent idiopathic scoliosis. Spine (Phila Pa 1976). 2010;35(2):194–198.
25. Good CR, Lenke LG, Bridwell KH, et al Can posterior-only surgery provide similar radiographic and clinical results as combined anterior (thoracotomy/thoracoabdominal)/posterior approaches for adult scoliosis? Spine (Phila Pa 1976). 2010;35(2):210–218.
26. Xie J, Wang Y, Zhao Z, et al Posterior vertebral column resection for correction of rigid spinal deformity curves greater than 100°. J Neurosurg Spine. 2012;17(6):540–551.
27. Mueller FJ, Gluch H. Cotrel-Dubousset instrumentation for the correction of adolescent idiopathic scoliosis. Long-term results with an unexpected high revision rate. Scoliosis. 2012;7(1):13.
28. Kepler CK, Wilkinson SM, Radcliff KE, et al Cost-utility analysis in spine care: a systematic review. Spine J. 2012;12(8):676–690.
29. Fishman LM, Sherman KJ, Groessl EJ. Serial case reporting yoga for idiopathic and degenerative scoliosis. Global Adv Health Med. 2014;3(5):16–21.
30. Cespedes A. How Often Should You Do Yoga a Week? http://http://www.livestrong.com
/article/474792-how-often-should-you-do-yoga-a-week/. Accessed January 30, 2017.
31. Modestini E. How Often Should You Practice Yoga? https://http://www.yogajournal.com
/practice/how-often-should-you-practice-yoga. Accessed November 21, 2016.
32. Ross A., Friedmann E, Bevans M, Tho S. Frequency of yoga practice predicts health: results of a national survey of yoga practitioners [published online ahead of print August 14, 2012]. Evid Based Complement Alternat Med. 2012;2012:983258. doi:10.1133/2012/983258.
33. Stasinaki AN, Gloumis G, Spengos K, et al Muscle strength, power, and morphologic adaptations after 6 weeks of compound vs. complex training in healthy men. J Strength Cond Res. 2015;29(9):2559–2569. doi:10.1519/JSC.0000000000000917.
34. Iyengar BKS. Light on Yoga. New York, NY: Schocken Books; 1966.
Keywords:Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.
Cobb angle; half-moon; isometric; scoliosis; side plank; tensegrity; yoga