J Manipulative Physiol Ther. 2001 (Jul); 24 (6): 385–393 ~ FULL TEXT
Charles A. Lantz, DC, PhD, Jasper Chen, DC
Marietta, GA, USA.
BACKGROUND: Chiropractors have long claimed to affect scoliotic curves, and case studies abound reporting on successful outcomes. No clinical trials exist, however, that evaluate chiropractic's effectiveness in the management of scoliotic curves.
OBJECTIVE: To assess the effectiveness of chiropractic intervention in the management of adolescent idiopathic scoliosis in curves less than 20 degrees.
DESIGN: Cohort time-series trial with all subjects electing chiropractic care. Entry-level Cobb angle was compared with postmanagement curve.
METHODS: Forty-two subjects completed the program of chiropractic intervention. Age range at entry was 6 to 12 years, and patients were included if their entry-level x-ray films revealed curves of 6 degrees to 20 degrees. Participants had adjustments performed for 1 year before follow-up. Full-spine osseous adjustments were the major form of intervention, but heel lifts and postural and lifestyle counseling were used as well.
RESULTS: There was no discernable effect on the severity of the curves as a function of age, initial curve severity, frequency of care, or attending physician.
CONCLUSION: Full-spine chiropractic adjustments with heel lifts and postural and lifestyle counseling are not effective in reducing the severity of scoliotic curves.
From the FULL TEXT Article:
The management of scoliosis poses formidable challenges to all health care practitioners. The changing attitude concerning the health risks of scoliosis has fostered a reevaluation of treatment methods and management programs for those with scoliosis.  The significant cost  and substantial risk  of surgery for scoliosis, along with earlier identification of patients with scoliosis through school screening programs, have led to a reconsideration of the role of conservative therapies in scoliosis management. [4-6]
Patients have been clearly shown to reject the medical position of the “3 Os”: observation, orthosis, and operation. In a group of patients who were assigned to wear braces, 50% refused to wear them, and 50% of those who were recommended for surgery refused that procedure as well.  Compliance with orthosis with medical bracing of scoliotic curves is a definite problem. [8-10] It would seem reasonable, then, to search for more acceptable conservative methods of treatment. Growing numbers of patients are seeking alternatives such as chiropractic,  and there has been a recent rise in popularity of electrical stimulation (ES), which has some problems of its own,  and the development of alternative “dynamic” bracing systems. [4, 13]
Consideration of conservative management of scoliosis requires a thorough understanding of its diagnosis and all aspects of its treatment. The effectiveness and significance of conservative management programs can be understood only when they are considered in light of the assumptions and tenets concerning scoliosis progression. Emphasis in this report is placed on the management of idiopathic scoliosis, since scoliosis associated with more apparent, genetically related conditions may not respond to conservative  or, for that matter, surgical  procedures. Since scoliosis in its milder forms is not recognized to cause any significant cardiovascular impairment or organic dysfunction, management is directed at the curve itself in an attempt to prevent it from progressing.  This is consistent with the trend in conservative management toward treating curves in their earlier stages of development, that is, milder curves, in an attempt to prevent future progression. 
Classically, conservative management of scoliosis has meant bracing and related orthotics administered under medical supervision. [6, 7, 18] Little attention has been paid to other forms of conservative management from a medical perspective except for a brief upsurge in ES. Virtually no credence has been given to chiropractic management programs as evidenced by the almost complete absence of discussion of chiropractic in medical literature on scoliosis despite a substantial number of patients with scoliosis who are receiving chiropractic care. 
Chiropractic has long claimed to offer a significant alternative to medical treatment for scoliosis, and many chiropractors claim substantial effects of their treatments on scoliotic spines. [19, 20] Chiropractic, however, like all other conservative methods, including many systems of bracing, has never been critically assessed by well-designed clinical trials.  Mawhiney claims that “a lumbar scoliosis of 10° or less should show up to 80% correction in 90 days. Predictable and assured”  but provides no evidence to support this contention. In light of what we know about the natural history of the condition, and the compelling logic of the biomechanical argument, such claims can appear reasonable. Whether this represents a significant improvement in the clinical picture, however, can only be assessed by large randomized controlled clinical trials. Simply reporting on cases that showed positive response [21-24] does not, as is shown in this study and elsewhere,  provide an assessment of the efficacy of the procedure. Few would contest the assertion that not every single curve would be expected to respond. Thus, a program of careful monitoring is required to track each case. The idea that chiropractic has a positive clinical impact on scoliosis bears further investigation for several reasons. First, the claim is made by some chiropractors that such corrections are, at least occasionally, possible and, in fact, actually occur (“Predictable and assured” ). Second, most orthopedists resign themselves to a very passive role in the early stages of the process (observation), and in the intermediate stages they seek to halt the progression, not correct the curve. At the stage of curve development to be studied in this project, medicine, until recently,  offered no active treatment programs.
The chiropractic clinical and popular literature is replete with overt claims and covert suggestions that chiropractic adjustments can correct scoliosis. [19-21, 30, 36, 37, 39] The study reported here represents the first attempt, however, at a significant clinical trial of the effectiveness of chiropractic in the management of scoliosis. We chose to focus exclusively on the magnitude of the curve, rather than attempt to assess the effect on quality of life and symptoms, such as pain, because patients seek out chiropractic (and other approaches) for the express purpose of correcting or reducing the severity of the curve. Overall, the advertising, screening, and recruiting programs were successful. Small curves (4° to 9°) were included in this study because this is common in chiropractic practices, and it was desirable to evaluate the entire spectrum of curves less than 25°. Curves less than 6° are those measured as secondary curves in double curve patterns. It was believed at the outset that, if any effects were to be observed, this range of curves and this age of subjects would be optimal. The procedures chosen for intervention consisted of full-spine systems with a major focus on osseous adjustive procedures. This consisted of Diversified and Gonstead systems, each primary attending chiropractor using one of these exclusively. The two are similar in that they are full-spine systems, they address the area of involvement (the curve itself), but they as well look to the regions below (especially the pelvis) and above (especially the cervical spine) the curves. Each session consisted of multiple adjustive procedures, with as many as 7 segments being adjusted.
The pattern of response of the curves reported here is what one might expect for the natural history of smaller curves although studies of curves less than 25° are lacking. It was reported for curves between 10° and 35° that, with no intervention, progression occurred in 20% of the subjects, improvement was seen in 25%, and no change was observed in 55% (with a filter of ±5°). [51, 52] The lack of any observable effect in the current study is disappointing in consideration of the care taken to select an optimally responsive population. In light of this, we think it is important to address certain issues and objections that might arise from the chiropractic community relative to the strengths and weaknesses of this study.
One of the major weaknesses of this study is the lack of a control group. This was by a design limitation, as a controlled clinical trial would have increased costs and time dramatically. It is also perhaps more appropriate to do a cohort study at this stage of development of research into this area. Arguably, with the study group selected, subjects would likely have gotten worse overall, and, therefore, chiropractic intervention prevented what would have been a much larger increase in curve. Although technically valid, this argument fails to acknowledge the design elements intended to optimize outcome in this study: younger subjects, smaller curves, frequent care. Only a controlled, randomized clinical trial can resolve this issue with proper design, analysis, and interpretation. A well-designed natural history study of this age group would be informative as well.
It will also, no doubt, be argued that it is not the purpose of chiropractic to straighten spines, and the important parameters, quality of life assessments, would have shown remarkable improvement. Or, for the chiropractic purist, the significant issue is whether there has been a removal of “nerve interference,” regardless of whether the spine was straightened.  We do not dispute these claims in this study and, indeed, it appeared that for those patients who had pain associated with their scoliosis there was a significant decrease in pain and improvement in quality of life. Only properly designed and controlled studies can resolve these issues.
It could also be argued that scoliosis is a 3-dimensional deformity and that other parameters of the curve might be affected. Although this is, indeed, true, it is our opinion that such changes, in the absence of changes in the coronal projection, are not terribly meaningful clinically or practically. We observed, for instance (results not shown), that vertebral rotation increased in some of the curves, especially those that progressed. Thus, derotation was not consistently observed in the curves, either.
Although some will argue that subjects were not followed long enough, it is our opinion that adequate time and attention were provided to see changes, if they were going to happen, with a mean of 41 visits (range, 6 to 124 visits). Several subjects continued for 2 or more years, and results were similar to those at 1 year.
There was a significant attrition rate among those who entered the study, most occurring within a few weeks of beginning care. We made no attempt to discover the reasons for their leaving. It was not, however, because they were disappointed with the results, because there was no follow-up x-ray examination before 4 to 6 months, depending on the severity of the curve. Chiropractic is a very personal experience and involves a lot of physical contact with the patient. It was our sense that some patients or their parents were uncomfortable with this situation. Gender issues played a definite role in several cases; some parents preferred female doctors for their female children.
One possible criticism of the study is the range of severity of curves studied. It could be argued that some “window of specificity” exists that is outside the range of inquiry (ie, 25°). Suppose, for instance, that 25° to 30° is a range in which some progression trigger is activated and progression ensues. Chiropractic intervention at this critical point could block progression. This and many other scenarios can be constructed, but such hypotheses will have to be assessed in the future to address these questions. For these future studies, some preliminary data would be most helpful and influential in initiating them.
The practical advantage of this design is that it addresses the claims made by chiropractors that early intervention can prevent scoliosis; thus the restriction of age and severity to early and mild is justified, though not all inclusive. Further study of patients with larger curves would be appropriate.
Great care was taken to develop an x-ray procedure that would allow for consistent placement at different examination times with the same subject and between subjects, and all studies were made with the same x-ray unit and technician. PA views were used to protect developing breast tissue. Although it has been shown  that PA views provide slightly larger curves than anterior-to-posterior (AP) views, the differences are minimal (2.4° for thoracic and 1.5° for lumbar curves; values for thoracolumbar curves were identical) and the rank order of curves was not affected. A “correction” value can be added or subtracted to obtain one or the other values. The PA view put the subjects facing the bucky, and to assure more consistent placement and posture of the head 2-point positioning was used with the nose and chin both in contact with the bucky. The hands were placed lightly at the sides of the bucky, again for stability and consistency, and the feet were rotated inward about 20°. Standard positioning with AP views does not provide adequate opportunity for stabilization and consistency of placement. One possible source of variability, however, was the growth of the subjects and the lack of exacting attention to uniformity in placement of the central ray along the longitudinal axis. The x-ray purist might argue that errors are skewed in the PA posture as opposed to AP, but this has never been shown to be the case, nor has any clinical significance been placed on it. There is, in fact, distortion in both positions, so we are at best talking about “relative” distortion. Further constructive discussion on this issue awaits some comparative measure of distortion and its clinical significance or impact. The additional risk of breast cancer by using the AP orientation must be justified by quantifiable benefits of the added x-ray clarity, even if such clarity can be shown to exist.
It will undoubtedly be argued by some that success would be assured if only some other procedure or technique system had been used in the intervention. A variant on this argument is that the practitioners (rather than the technique system used) were somehow deficient, presumably because of lack of experience. One report identified about 100 named technique systems in chiropractic,  many with only a handful of practitioners. The procedures used in this study, however, are practiced by the vast majority of chiropractors in the United States and abroad, according to a survey of chiropractors.  It is possible that some of these practices have a pronounced effect on scoliosis, but the promoters of these technique systems are challenged to provide credible data to support any assertions they might make regarding the effectiveness of their systems. There are, for example, case reports of correction of scoliosis by procedures that adjust only the atlas, the superior-most vertebra of the spine. [22, 24] About 15 systems promote this “upper cervical” approach to chiropractic. Case studies [21-24] are hardly the evidence necessary to support assertions of effectiveness of a procedure. Indeed, in the present study, subjects were observed to improve significantly (up to 13° reduction), and sometimes, especially in smaller curves, to return to almost completely straight spines (100% correction) or to reverse the curve (more than 100% change). Overall, however, the changes were clinically insignificant (less than 1° correction). Again, the clinician will argue, “But it is significant for the patient that improves.” To this we would reply, “Yes, and all the more so for those who would have progressed.” The above line of reasoning represents an intriguing challenge to the design of the study and tends toward the heart of chiropractic management approaches. Are there more and less appropriate techniques for specific conditions, or what is the link between the pattern, frequency, or degree of the subluxation and the changes in patient welfare (and in this case, structure)? Is this the secret wealth of chiropractic or yet another empty wish? Only further research will be able to resolve this issue.
On the basis of the results presented here, it must be concluded that full-spine osseous adjustments, according to Diversified and Gonstead technique systems, supplemented with heel lifts and postural and lifestyle counseling, were not effective in the correction of curvature in scoliotic spines, as determined by PA projections on plane films in children between the ages of 6 and 17 years with curves less than 25°. The lack of a natural history of scoliosis less than 25° precludes determination of whether curves were prevented from worsening. The effect of chiropractic care on quality of life or “nerve interference” was not determined in this study.
We thank John Barnard, Romier Visperas, Brian Gatterman, DC, Trent Bachman, DC, Sandra Coleman, and the children and their parents without whom this study would not have been possible.