Content uploaded by Stefano Negrini
Author content
All content in this area was uploaded by Stefano Negrini
Content may be subject to copyright.
Exercises reduce the progression rate of adolescent idiopathic scoliosis:
Results of a comprehensive systematic review of the literature
S. NEGRINI, C. FUSCO, S. MINOZZI, S. ATANASIO, F. ZAINA & M. ROMANO
ISICO (Italian Scientific Spine Institute), Milan, Italy
Abstract
Background. A previously published systematic review (Ped.Rehab.2003 – DARE 2004) documented the existence of the
evidence of level 2a (Oxford EBM Centre) on the efficacy of specific exercises to reduce the progression of AIS (Adolescent
Idiopathic Scoliosis).
Aim. To confirm whether the indication for treatment with specific exercises for AIS has changed in recent years.
Study design . Systematic review.
Methods. A bibliographic search with strict inclusion criteria (patients treated exclusively with exercises, outcome Cobb
degrees, all study designs) was performed on the main electronic databases and through extensive manual searching. We
retrieved 19 studies, including one RCT and eight controlled studies; 12 studies were prospective. A methodological and
clinical evaluation was performed.
Results. The 19 papers considered included 1654 treated patients and 688 controls. The highest-quality study
(RCT) compared two groups of 40 patients, showing an improvement of curvature in all treated patients after six months.
We found three papers on Scoliosis Intensive Rehabilitation (Schroth), five on extrinsic autocorrection-based methods
(Schroth, side-shift), four on intrinsic autocorrection-based approaches (Lyon and SEAS) and five with no autocorrection
(three asymmetric, two symmetric exercises). Apart from one (no autocorrection, symmetric exercises, very low
methodological quality), all studies confirmed the efficacy of exercises in reducing the progression rate (mainly in early
puberty) and/or improving the Cobb angles (around the end of growth). Exercises were also shown to be effective in reducing
brace prescription.
Conclusion. In five years, eight more papers have been published to the indexed literature coming from
throughout the world (Asia, the US, Eastern Europe) and proving that interest in exercises is not exclusive to Western
Europe. This systematic review confirms and strengthens the previous ones. The actual evidence on exercises for AIS is of
level 1b.
Keywords: Physical exercises, adolescent idiopathic scoliosis, conservative treatment, physiotherapy, rehabilitation
Introduction
Various types of treatments for AIS (Adolescent
Idiopathic Scoliosis), whether conservative or surgi-
cal, have been reported. The majority of adolescents
with AIS have been treated with conservative care
that included bracing, simple observation and/or
physical exercises (PEs) [1]. PEs for the treatment of
AIS have been used since 500 BC, when Hippocrates
[2], followed by Galenus [3], introduced their usage
as means to maintain the flexibility of the chest wall.
During the past centuries there was a considerable
flowering of different approaches to PEs, but only at
the beginning of the previous century, mainly in
Germany with Klapp and Von Niederho¨ fer, was it
possible to verify the first methods through deep
scientific observation [4]. During the same period
Katharina Schroth described her method [5]. Later,
in many parts of Europe, authors described different
methods: Between 1930 and 1950 the ‘IOP’ method
was introduced in Italy, and the ‘Psoas’ method was
produced in the Soviet Union [4]. The ‘Lyon’
method [6,7] and that of Me´zie`res [8] were
described in France during the 1960s. Later,
Souchard derived its treatment from Me´zie`res [9],
and in Poland Dobosiewics proposed its approach
Correspondence: Stefano Negrini, Direttore, Scientifico ISICO (Istituto Scientifico Italiano Colonna vertebrale), Milano, Italy. E-mail: stefano.negrini@isico.it
Disability and Rehabilitation, 2008; 30(10): 772 – 785
ISSN 0963-8288 print/ISSN 1464-5165 online ª 2008 Informa UK Ltd.
DOI: 10.1080/09638280801889568
Disabil Rehabil Downloaded from informahealthcare.com by Dr Stefano Negrini on 03/30/11
For personal use only.
[10], while Min Mehta introduced the ‘side-shift
therapy’ [11] to arrive at the actual ‘Scientific
Exercises Approach to Scoliosis (SEAS),’ again
created in Italy [12].
Apart from each method, PEs were subjected
during the years to various fates [13,14], and today
the current evidence regarding the effectiveness of
PEs and other conservative treatments for AIS
remains insufficient [14 – 16]. At the moment there
is certainty only in regard to the ineffectiveness of
electrical stimulation [17]. Despite the fact the
rationale for the choice of which conservative
treatment should be used is unclear and the
effectiveness of exercises has not been proved, in
many areas of the world patients are expected to
continue treatments that have an impact on their
quality of life [14]. We are perfectly aware that
exercises also achieve other important objectives
[18,19], such as increased neuromotor control and
spinal stability, biomechanical reduction of postural
collapse and increased respiratory function. More-
over, the effectiveness of PEs in AIS patients has
already been shown to improve respiratory function
[20,21], strength [22] and postural balance [23]. But
a key factor for the scoliosis treating community is to
understand whether exercises are truly capable of a
positive influence on the deformity [18], and this is
the main reason we have conducted, after five years,
an update on our previous systematic review [14]
about PEs as a treatment to avoid the progression of
AIS and ultimately prevent the use of bracing. The
goal of this paper is to systematically search the
literature for news on the effectiveness of PEs as
means to eventually modify, and therefore conclude,
our review.
Methods
We followed exactly the same methodology of our
previous review [14] so as to add our actual results to
the previous ones. We performed an extensive search
through all the relevant databases: Medline, Embase,
Cinhal, PEDro and Cochrane Library. We searched
papers in every available language covering the
period from the month we concluded our previous
search (December 2002) up to July 2007. Again, we
used the Mesh terms ‘scoliosis AND (‘‘exercise
therapy OR ‘rehabilitation’’)’, and in free text we
used ‘idiopathic scoliosis’ followed by the operator
‘AND’ and the terms ‘exercise,’ ‘exercises,’ ‘sports,’
‘sport,’ ‘rehabilitation’ and ‘physiotherapy.’ No
restriction was applied in regard to language. We
checked the reference list of the retrieved articles and
also performed a manual search of the journals
listed in Table I. The inclusion criteria remained the
same:
. Patients: Diagnosis of AIS by a specialist,
confirmed through X-rays; we focused on
patients in growth years (up to Risser 5) and
accepted three studies in which most of the
patients were growing, but some older ones
were also included;
. Experimental intervention: Patients treated
exclusively with PEs, without any other asso-
ciated intervention;
. Control group: Any kind of patients, either
observed or treated;
. Outcome measures: Only Cobb degrees; re-
sults could be reported in absolute terms or as
percentage of patients improved/worsened;
. Study design: Any study design.
For the methodological evaluation of the studies
we considered the following parameters (Table II)
[14]: Controlled study, random allocation to experi-
mental and control intervention, prospective versus
retrospective study, recruitment modality described,
patient characteristics described, intervention de-
scribed, blinded assessment of outcomes, identifica-
tion of possible confounding factors, statistical
control for the confounding factor.
The results are herein reported in methodological
and clinical terms. In the clinical section we will look
for the duration of the studies and number of patients
included, but also the characteristics of patients as
well as the PE techniques used. Accordingly, the
clinical presentation will be divided for:
. Methodology (randomized controlled trial –
RCT – versus others): Obviously good meth-
odology papers can be trusted more than the
others;
. Type of rehabilitation (Scoliosis Intensive
Rehabilitation – SIR – versus outpatient
rehabilitation exercises – [OR]): Inpatient
treatment is totally different in many respect
(possible efficacy but also costs) from out-
patient one;
Table I. Journals in which the manual search was performed.
Journal Years searched Language
Annales de Kine´sithe´rapie 2002 – 2007 French
Kine´sithe´rapie Scientifique 2002 – 2007 French
Re´sonances Europe´ennes Du Rachis 2002 – 2007 French
Cahiers de Kine´sithe´rapie 2002 – 2007 French
Ginnastica Medica, Medicina
Fisica e Riabilitazione
2002 – 2007 Italian
Chinesiologia Scientifica 2002 – 2007 Italian
Atti Gis, Giornate di Patologia
Vertebrale
2002 – 2007 Italian
European Medical Physiology 2002 – 2007 English
European Spine Journal 2002 – 2007 English
Exercises reduce the progression rate of AIS: Review of literature 773
Disabil Rehabil Downloaded from informahealthcare.com by Dr Stefano Negrini on 03/30/11
For personal use only.
Table II. Methodological quality of retrieved studies.
Study
Controlled
study
Random
allocation
Prospective
study
Allocation
criteria
described
Recruitment
modality
described
Patient
characteristics
described
Intervention
described
Blinded
assessment of
outcomes
Confounding
factors
Author Year Ref Identification Statistical control
Wan 2005 31 Y Y Y Y: random Y Y Y N N N
Weiss 1997 14 N N Y Na N Y N N N N
Weiss 1992 40 N N N Na N Y Y N N N
Weiss 2003 16 Y N Y N N Y Y N Y N
Rigo 1991 17 N N N Na N Y Y N N N
den Boer 1999 19 Hc N Y Na Y Y Y N Y N
Mamyama 2002 20 N Na Y Na N Y Y N N N
Maruyama 2003 21 N Na N Na N Y Y N N N
Otman 2005 18 N Na Y Na Y Y Y N N N
Mollon 1986 27 Y N N N N Y N N Y Y
Ferraro 1998 29 N N Y Na Y Y Y N N N
Duconge´ 2002 28 Y N N N N Y N N N N
Negrini 2006 30 Y N Y Y: patient’s decision Y Y Y N N N
Mooney 2000 23 N N Y Na N Y Y N N N
Durmala 2003 26 N Na Y N N Y Y N N N
McIntire 2006 24 N Na Y Na N Y Y N N N
Stone 1979 22 Hc N Y Na Y Y Y N N N
Klisic 1985 32 Y N N Y: patient’s decision N N N N N N
Negrini 2006 31 Y N Y Y: patient’s decision Y Y Y N N N
Y, yes; N, no; NA, not applicable; HC, historical control.
774 S. Negrini et al.
Disabil Rehabil Downloaded from informahealthcare.com by Dr Stefano Negrini on 03/30/11
For personal use only.
. Type of autocorrection (AC) exercises
proposed (extrinsic: obtained mainly through
the arms and/or trunk muscles; or intrinsic:
focused mainly on contraction of paravertebral
spine muscles; or no AC); autocorrection
has been proposed by SOSORT Consensus
[18] as the key to a proper rehabilitation
protocol;
. In cases where the AC (i.e., asymmetrical by
definition) [18] was not used, we will differ-
entiate between symmetric and asymmetric
exercises.
Results
The electronic search allowed us to find 192
articles, but through the manual search we did
not find relevant papers. In this way the papers
considered, including those coming from the
previous review, [14] amounted to 616. Reading
the titles and abstracts we excluded 586 papers,
while we searched 30 full texts for further evalua-
tion as probably relevant. They were all retrieved:
Ten were excluded because they did not fulfil all
the inclusion criteria, and the final review included
19 papers.
Characteristics of the included studies
The 19 relevant studies (Tables III, IV and V)
encompassed 2342 AIS patients, 1654 of whom had
been treated exclusively with PEs for an average time
of 23 + 18 months. The age considered in the
studies was on average 13.4 + 2.8 years, and the
scoliosis curvature was 23.6 + 8.98 Cobb. The
control groups included for observation [24 – 28] or
alternative treatments (other kinds of exercises
[29 – 31], and in one case bracing [11]). We
excluded any paper that considered results other
than Cobb degrees: Two studies [31,32] did not
report clinically on the worsening or improvement of
the patients, while three [25,28,33] did not include
the final curvatures obtained. One study [30]
reported on the usefulness of exercises in enhancing
the efficacy of bracing [34,35].
Given the heterogeneity of the studies and their
weak methodologies, we did not attempt a real meta-
analysis. Nevertheless, a general purview of all
studies was possible. First, all groups treated with
exercises had better clinical results than the five
observed control groups (Figure 1). Generally, it
seems that at the beginning of adolescence and the
start of puberty, when the risks of progression are
Figure 1. The clinical results (patients improved, unchanged or worsened) reported in all exercises groups in the studies considered in
this review are better than those reported in the observational groups. All the observational groups (obs) are listed under the black line,
all the exercise groups (ex/con) over this line. The exercise groups include also a control group (con) that performed usual
physiotherapy.
Exercises reduce the progression rate of AIS: Review of literature 775
Disabil Rehabil Downloaded from informahealthcare.com by Dr Stefano Negrini on 03/30/11
For personal use only.
higher, patients with lower-degree curvatures are
usually treated, and the results show a slow progres-
sion and rarely a decrease in curvature, while in
patients of older ages success is achieved more easily
and in greater curvatures (Figure 2). In any case, the
studies with the biggest numbers of patients tend to
show good results (stability) but such results are
generally not as significant as those with reduced
samples (Figure 3).
Methodological quality
The results of quality assessments of the included
studies are shown in Table II. The quality of the
studies was found to be weak. Only one study was a
randomized controlled trial [31], but its methodolo-
gical quality was poor: The method of sequence
generation and of allocation concealment was not
described, the outcome assessor was not blinded,
and none was lost at follow-up. Six studies were
subjected to concurrent controls [25 – 30], and two
had historical controls [11,24]. Only three out of the
six studies with concurrent controls were prospective
[25,29,30]. In the two studies with historical controls
the experimental group was evaluated prospectively.
The allocation criteria (other than random criteria)
were described only in three [28 – 30] of the six
controlled studies, and were based on the patient’s
decision. The other 12 included studies that were
uncontrolled and involved only one evaluation of the
outcome measure before and after treatment. Seven
of these studies were prospective [22,32,33,36 – 39].
The recruitment modality was described in seven of
the 19 studies included [11,24,29 – 32,37]. The
characteristics of patients were described in all the
studies, whereas the description of the intervention
was reported in 16 of 19 studies. Only three studies
attempted to identify the possible confounding
factors [11,25,26], and only one [26] adjusted the
results accordingly.
Clinical results
RCT (Randomized controlled trial)
Wan [31], in 2005, published the first RCT about
PEs for the treatment of AIS in 80 Chinese
patients (40 per group) of 15 + 4 years of age
and 24 + 128 of curvature: All patients received
electrostimulation on lateral body surface as well as
traction and postural training; and patients in the
treatment group underwent specific asymmetric
Figure 2. Results in terms of Cobb degrees of the studies included in this review according to the age of the treated patients. In this graph, the
thick black lines represent the treated patients, the thick dotted lines (Wan 2005 and Negrini 2006) the control exercise groups, the thin
dotted lines (Mollon 1986 and Duconge` 2002) the control observational groups, and the thin dashed lines (den Boer 1999) the only control
braced group.
776 S. Negrini et al.
Disabil Rehabil Downloaded from informahealthcare.com by Dr Stefano Negrini on 03/30/11
For personal use only.
strengthening PEs once daily for correction of
scoliosis. With six months of treatment the changes
of Cobb angle in the thoracic and lumbar segments
were significant, but in the PE group the improve-
ment (158) was greater than in the control group
(78).
Figure 3. The papers with the biggest number of patients have results of stability of the curves (+58).
Table III. Characteristics of the included studies: Populations and follow-up. The study by Weiss et al. has been divided into two age groups
according to the original study.
Study Population Duration of treatment
Number of patients Age Cobb degrees Months
Author Year Ref Total Exercises Controls Average SD Range Average SD Range Average SD Range
Wan 2005 36 80 40 40 15 4 6
Weiss 1992 15 107 107 21.6 10.9 – 48.8 43 10 – 114 1.5 1 – 1.5
Weiss 1997 14 181 181 12.7 27 33
Weiss 2003 25a 94 30 64 10 4 – 11 21 5 – 52 35 23
Weiss 2003 25b 102 59 43 13 12 – 14 29.5 5 – 68 34 37
Rigo 1991 17 43 43 12 19.5 19.5 3 – ?
den Boer 1999 19 164 44 120 13.6 10 – 15 26 20 – 32 26 4 – ?
Mamyama 2002 20 69 69 16.3 11 – 27 31.5 13 – 74 50 14 – 132
Maruyama 2003 21 53 53 16.3 13 – 27 33.3 20 – 74 41 12 – 132
Otman 2005 18 50 50 14.1 11 – 17 26.1 20 – 35 12
Mollon 1986 27 210 160 50 10.8 10 – 15 16 53
Ferraro 1998 29 34 34 11.6 8 – 14 14.9 10 – 24 24 7 – 51
Duconge´ 2002 28 591 422 169 10.1 7 – 16 15.6 55
Negrini 2006 30 48 23 25 12.4 15.1 12
Mooney 2000 23 12 12 13.1 11 – 16 33.5 20 – 60 4
Durmala 2003 26 136 136 6 – 18 12
McIntire 2006 24 9 9 14 1.7 29 6 20 – 37 4
Stone 1979 22 99 42 57 12 – 15 10 4 – 22 12 9 – 15
Klisic 1985 32 150 100 50 11 5 – 15 14 36 12 – 84
Negrini 2006 31 110 40 70 13.4 10 – 15 30.9 5
Exercises reduce the progression rate of AIS: Review of literature 777
Disabil Rehabil Downloaded from informahealthcare.com by Dr Stefano Negrini on 03/30/11
For personal use only.
Table IV. Characteristics of the included studies: Exercises performed and compliance.
Study Exercises
ComplianceAuthor Year Ref Type Aims Characteristics
Wan 2005 13 Gymnastic exercise
and postural
training (Na)
Correction of essential
S-shaped scoliosis and
maintaining symmetrical
posture during normal
activities
Once daily
Weiss 1997 14 Schroth method (E) actively straightening,
auto-correction
intensive in-patient exercise
programme
Weiss 1992 15 Schroth method (E) actively straightening,
auto-correction
intensive in-patient exercise
programme
Weiss 2003 16 Schroth method (E) actively straightening,
auto-correction
intensive in-patient exercise
programme
Rigo 1991 41 Schroth method (E) actively straightening,
auto-correction
three times a week with
physio
Den Boer 1999 11 Side shift therapy (E) side shift, posture instruction from physio,
side-shift in daily living
95%
Mamyama 2002 38 Side shift therapy (E) Lateral shift-trunk during
standing and sitting
Not reported Not reported
Maruyama 2003 42 Side shift therapy (E) Lateral shift-trunk during
standing and sitting
Otman 2005 18 Scroth method (E) Active extension, active-
correction, realignment
of trunk
Outpatient rehabilitation
for the first 6 weeks (4
hours day for 5 days) and
later same program at
home
68 patients
recruited,
18 not
compliant
excluded from
the analysis
Mollon 1986 26 Lyon method (I) posture control,
strengthening, balance
twice a week with physio
and other times at home
75%
Ferraro 1998 29 Many methods (I) active postural correction twice a week with physio
and other times at home
see text
Duconge´ 2002 27 Lyon method (I) posture control,
strengthening, balance
twice a week with physio
and other times at home
71%
Negrini 2006 29 SEAS.02 (I) Active-self correction
learning
Individually adapted
exercises at a super-
specialized structure
(1.5 h session every 2 – 3
months), prosecution at
home
Not reported
Mooney 2000 23 MedX Rotary Torso
Machine (Na)
strengthening Daily
McIntire 2006 24 Trunk rotational
strength training
(Na)
Increase trunk strength,
stabilize or decrease
curve size
2 training session per week
(tot 32)
Not reported
Durmala 2003 33 Asymmetric
mobilization of the
trunk (Na)
Stone 1979 22 Milwaukee method
(Ns)
mobilization, strengthening,
posture
instructions from physios,
exercises performed at
home
about 50%
Klisic 1985 32 (Ns)
Negrini 2006 31 SEAS.02 (Ns) Active-self correction
learning
Individually adapted
exercises at a super-
specialized structure
(1.5 h session every 2 – 3
months), prosecution at
home
Not reported
Abbreviations: E: extrinsic autocorrection; I: Intrinsic autocorrection; Na: No autocorrection, asymmetric exercises; Ns: No autocorrection,
symmetric exercises.
778 S. Negr ini et al.
Disabil Rehabil Downloaded from informahealthcare.com by Dr Stefano Negrini on 03/30/11
For personal use only.
Other studies
SIR (Scoliosis intensive rehabilitation)
This sub-group includes three studies, 484 AIS
patients 14.3 + 5 years old, with 30 + 98 Cobb, in
which 377 had been treated for 26 + 16 months. All
the studies reported on this kind of approach have
been proposed by Weiss [25,36,40], using the
intensive physiotherapy protocol originally proposed
by Katharina Schroth (4 – 6 weeks of treatment, 6 – 8
hours per day, then at home for 90 min per day).
The Schroth method (Figure 4) is based on
sensorimotor and kinaesthetic principles. Exercises
comprise the correction of scoliotic posture (through
elongation, realignment of trunk segments, position-
ing of the arms and muscle contraction), and the use
of specific breathing patterns with the help of
proprioceptive and esteroceptive stimulation and
mirror control.
The authors first proposed two case series: In 1992
[40] they proved in 107 patients of various ages
(ranging from 10.9 – 48.8 years) with curvatures of
438 (ranging from 10 – 1148) the efficacy of SIR in
the short term (4 – 6 weeks), with an improvement in
44% and a worsening in 3%; in 1997 [36] 181
adolescents (12.7 years) with 278 curvatures did not
progress in 33 months (end of treatment: 298 – 25%
worsened, 18% improved). Finally, in 2003 [25] the
authors proposed a prospective cohort controlled
study in two sub-groups matched for sex and age
(first subset, ten years of age 218 curves; second one,
13 years old, 29.58 curves). In both SIR groups the
authors found in 35 months of treatment better
results than in control groups: The youngest had
53% improvement versus 29% in controls, the oldest
70% versus 44%, respectively.
OR (Outpatient rehabilitation exercises)
Extrinsic autocorrection. This sub-group includes five
studies, 379 AIS patients 14.4 + 1.8 years old, with
27 + 58 Cobb; 259 have been treated for 32 + 16
months. Two papers used the Schroth method as an
outpatient approach. In the retrospective study by
Rigo [41], 43 patients 12 years old with 19.58 curves
exercised twice a week following a two-hour program
Table V. Results of the retrieved studies.
Study
Groups
8Cobb Clinical change
Author Year Ref Start End P Variation considered Improved Unchanged Worsened P
Wan 2005 13 Specific exercises 25 10
Postural exercises 24 17
Weiss 1992 15 Exercise 43 39 5 44% 53% 3%
Weiss 1997 14 Exercise 27 29 5 18% 57% 25%
Weiss 2003 16a Exercise 21 53% 47%
Observation 5–308 29% 71%
Weiss 2003 16b Exercise 1 29 70% 31%
Exercise 2 42 81% 19%
Observation 5–308 44% 56%
Rigo 1991 17 Exercise 19 17 5 12% 44% 44%
den Boer 1999 19 Exercise 26 28 NS
Brace 27 25
Mamyama 2002 20 Exercise 31 30 5 22% 64% 14%
Maruyama 2003 21 Exercise 33 32
Otman 2005 18 Exercise 26 18 5 100% 0% 0%
Mollon 1986 27 Exercise 17 18 3 63% 4% 34%
Observation 13 23 20% 5% 75%
Ferraro 1998 29 Exercise 15
Duconge´ 2002 28 Exercise 17 21 3 58% 42%
Observation 12 25 23% 77%
Negrini 2006 30 Exercise 15 12 5 29% 68% 3%
Usual physiotherapy 15 15 5% 82% 13%
Mooney 2000 22 Exercise 33 27 5 42% 50% 8%
McIntire 2006 39 Exercise 29 24 5 44% 33% 22%
Durmala 2003 26 Exercise: double curves 33% 27% 39%
Exercise: single curves 39% 26% 35%
Stone 1979 24 Exercise 10 8 4 21% 74% 5%
Klisic 1985 28 Exercise 15 58% 5% 37%
Observation 13 26% 10% 64%
Negrini 2006 30 Exercise 30 25 5 58% 41% 1%
Usual physiotherapy 31 28 46% 44% 10%
Exerc, exercises; Ctrl, controls. Ref. No. 16 (Weiss et al.) has been divided, according to the original study, into two different age groups (a)
and (b); the group (b) included two sub-groups (1) and (2): This presented the worst curvatures.
Exercises reduce the progression rate of AIS: Review of literature 779
Disabil Rehabil Downloaded from informahealthcare.com by Dr Stefano Negrini on 03/30/11
For personal use only.
for a minimum of three months: 11.6% worsened
and 44.2% improved. Recently, Otman [37] fol-
lowed prospectively a series of 50 adolescents 14.1
years old (ranging in age from 11 – 17) as outpatients
for the first six weeks (four hours per five days a
week) who continued exercising at home for 90 min
per day: During the treatment he observed a
progressive improvement of Cobb degrees from an
average angle of 26.1 – 17.88 at one year, when all
patients showed improvement.
Another method that is very common (and has
been studied by several authors) is the ‘side-shift
therapy’ (Figure 5) originally proposed by Min
Mehta. This is a kind autocorrection, first through
a lateral shift of the trunk applied as an exercise and
then in the activities of daily living. The oldest study
Figure 4. Example of a Schroth therapy exercise (extrinsic autocorrection). Permission has been given by Dr. Weiss to reproduce this figure.
Figure 5. Side-shift exercise (extrinsic autocorrection). A: neutral position; B: end of side-shift moevement. (Courtesy Dr. T. Maruyama)
780 S. Negr ini et al.
Disabil Rehabil Downloaded from informahealthcare.com by Dr Stefano Negrini on 03/30/11
For personal use only.
has been proposed by den Boer [11], who compared
prospectively PEs (44 patients) with bracing (120
adolescents) in adolescents 13.6 years old with
20 – 328 Cobb curvatures. The authors found no
statistically significant differences or the intention to
treat or determine efficacy between the two groups.
Mamiyama, in 2002 [38], prospectively followed a
series of 69 patients 16.3 years old with 31.58
curvatures who started exercising after skeletal
maturity (Risser grade IV or V, postmenarche after
more than two years): After an average follow-up
period of 4.2 years the authors observed that 22% of
the curves had progressed while 14% improved (4%
of 108 or more). One year later Maruyama [42]
retrospectively confirmed these results in a sample of
53 patients of similar clinical characteristics treated
for 41 months.
Intrinsic autocorrection. This sub-group includes four
studies, 883 AIS patients 11.2 + 1.0 years old, with
15 + 0.58 Cobb, in which 259 had been treated for
36 + 21 months. This intrinsic AC approach is based
on the former Lyon School proposals, which were
based on the auto-elongation exercise [6,7]
(Figure 6). According to the new knowledge in the
literature, this have been reinterpreted in terms of
three-dimensional AC in the SEAS (Scientific
Exercises Approach to Scoliosis) by ISICO (Italian
Scientific Spine Institute) [12] (Figure 7).
Figure 7. Active autocorrection according to the SEAS protocol proposed by the ISICO School (intrinsic autocorrection). Copyright
permission has been granted by ISICO to reproduce this figure.
Figure 6. Auto-elongation according to the Lyon School (intrinsic
autocorrection). Copyright permission has been granted by ISICO
to reproduce this figure.
Exercises reduce the progression rate of AIS: Review of literature 781
Disabil Rehabil Downloaded from informahealthcare.com by Dr Stefano Negrini on 03/30/11
For personal use only.
The oldest retrospective controlled study has been
proposed by Mollon and Rodot [26], which studied
the entire growth period of 210 AIS patients 10.1
years old recruited throughout Europe: These 50
controls and 160 patients had been treated with PEs
consisting of posture control, strengthening and
balance training according to the Lyon method
[6,7], based on the neuromotor control to be in-
creased though external input, mainly propriocep-
tive. At the end of treatment (four years) the authors
found statistically significant differences, with 62.5%
of treated patients improved (44% worsened), while
in the control group these percentages were 20% and
75%, respectively. More than ten years later,
Duconge´ [27] confirmed these results in a similar
sample with the same methods: 42% of the 422
treated patients worsened versus 77% of the 169
controls. A similar treatment has been applied in Italy
by Ferraro [32], who treated 34 outpatients twice a
week (plus personalized home sessions). When com-
pared to the minimal compliance (5 10 min/day), the
maximal participation (430 min/day) slowed down
or even halted the progression of scoliosis (difference
between the groups: 98).
More recently, Negrini [29] compared to usual
physiotherapy in a prospective study the SEAS
protocol, which is based on active three-dimensional
self-correction (Figure 7) and aims at improved
spinal stability, developing balance reactions and
preserving/improving the physiological sagittal spinal
curvatures. In this one-year prospective controlled
study of 48 patients 12.4 years old, with 158 curves,
the number of braced patients (failure of treatment)
was statistically significantly reduced by specific
exercises (4.3% versus 20%), while SEAS caused
28.9% improvements versus 5% in controls.
No autocorrection, asymmetric exercises. Some authors
did not apply the autocorrection principles but
instead used asymmetric exercises according to
different theories. This sub-group includes three
studies. Durmala [33] followed prospectively (for 12
months) 136 AIS patients ranging in age from six to
18 years, treated with asymmetric trunk mobilization
and reported a 31 – 39% decrease in Cobb angle
according to the segment of the spine considered.
Two pilot studies proposed PEs performed with the
aid of a rehabilitation device, the MedX Rotatory
Torso Machine for trunk rotation strength training,
starting with the idea that there is increased activity
on the convex side of scoliosis, particularly in
patients with progressive curvatures. Mooney [22]
performed a pilot study and obtained in a period of
four months, in 12 patients 13.1 years old with 33.58
curvatures, an improvement in 41.6%, while only
one worsened. Six years later another pilot study by
McIntire [39], who proposed, to nine patients aged
14 years with 298 curvatures, two training sessions
per week for four months and reported a statistically
significant reduction of 58.
No autocorrection, sym metric exercises. Finally, there is
the group of authors who neither applied the
autocorrection principles nor used asymmetric ex-
ercises. This sub-group includes three studies, the
oldest ones (and the only one negative) and one of
the last, which relate to a particular possible aim of
PEs, being increased brace correction. In 1979,
Stone [24] proposed to 42 patients ranging in age
from 12 – 15 years with 108 curvatures (range 4 – 22),
a 12-month home program of mobilization, strength-
ening and posture control according to the Milwau-
kee method, but obtained no differences between the
prospective PE group and 57 retrospective controls.
Contrastingly, in 1985 Klisic and Nikolic [28],
through PEs (no details of protocol given) obtained
an improvement in 58% of 100 patients 11 years of
age with 148 curvatures, while 37% worsened versus
26% and 64%, respectively, in 50 controls. Finally,
one paper explored the usefulness of PEs in
preparation for bracing [30]: The SEAS protocol of
general and localized mobilization demonstrated its
efficacy in increasing the brace correction at five
months in 40 patients versus 70 control patients 13.4
years old with curvature of 30.98. Of these, there was
a 58% improvement rate and 1.5% worsening in the
PE-treated versus 45.8% and 10.3%, respectively, in
controls.
Discussion
This review adds a lot of new data to the previous
one we performed five years ago [14]. In this period
an RCT [31] and seven more papers of different
methodological quality have been published
[29,30,33,37 – 39,42]. All results are consistent with
the previous ones [14,15,43], and confirm a possible
efficacy of PEs in contrasting the progression of AIS.
The evidence is not of the highest level (1b,
according to the Oxford Centre for Evidence-Based
Medicine) [44], but it is better than the previous
existing evidence of level 2a [14].
The RCT is the strongest research design on the
basis of which to draw valid conclusions regarding
the effectiveness of a therapeutic intervention be-
cause, if well conducted, it minimizes the risk of bias.
Nonetheless there are many clinical settings in which
RCTs are difficult, impractical or unethical. In such
situations a controlled non-randomized study, an
observational controlled study or an uncontrolled
study could constitute a valid alternative, providing
that confounding factors and sources of bias are
carefully analyzed [45]. Rehabilitation is among the
medical fields in which a researcher attempting to
782 S. Negr ini et al.
Disabil Rehabil Downloaded from informahealthcare.com by Dr Stefano Negrini on 03/30/11
For personal use only.
conduct an RCT is most likely to encounter
difficulties [46]: It is often difficult to collect a
homogenous patient sample large enough to obtain
adequate power of the study. Additionally, it is often
difficult to find a suitable placebo intervention, and it
is sometimes impossible for ethical and practical
reasons to include a ‘no intervention control group’.
Moreover, it may often be impossible to distinguish
between the specific effect of the intervention (PEs,
physical therapies or other) and the therapeutic effect
(psychological) of the patient-therapist relationship.
Equally, the specific effect of the intervention could
be modified by the therapist’s expertise and faith in
the given technique. Finally, it is almost always
impossible to establish a double-blind condition.
Nonetheless, it is possible to evaluate the effective-
ness of rehabilitative intervention through good-
quality studies. There are, in fact, many published
studies that try to overcome the difficulties associated
with the field of rehabilitation. The Cochrane
Library, for example, contains many systematic
reviews on various rehabilitative interventions, which
include RCTs of acceptable quality that try to
overcome said difficulties [47].
Only one of the retrieved studies regarding the
effectiveness of PEs in AIS was randomized. Twelve
of the included studies were uncontrolled and
involved only one evaluation of the outcome measure
before and after treatment. This kind of design
produces results that are impossible to interpret,
because one cannot reasonably conclude that the
improvement observed was causally determined by
the intervention: The positive change could have
occurred naturally or might have been the result of
other aspects of therapy being conducted contem-
poraneously [45]. Six studies were controlled with
concurrent controls, and two had historical controls.
Moreover, the controlled and uncontrolled studies
gathered in the authors’ search failed to meet some
methodologic criteria for observational studies. It is
therefore impossible, on the basis of the data
contained in these studies, to draw strong conclu-
sions on the effectiveness of PEs in AIS. Further
randomized controlled trials or properly conducted
observational prospective studies with adjustments
for confounding factors should be realized.
Another topic to be considered when deciding
whether a treatment can be applied is the kind of
adverse effect: In all these studies (nor in any other
we know) the adverse effects of the exercises
performed have been documented. Eventually, the
only problem is in terms of dropouts and compliance
with treatment over the long term; moreover, these
aspects have not been well documented. In any case,
this is a key factor for rehabilitation studies that has
not been thoroughly discussed in most of the studies
reported in this review.
According to the previous review [14], it was
possible to imagine that PE as a form of treatment
was confined mainly to Western Europe, because
papers arrived mainly from Germany [25,36,40] and
France [26,27], but also from Holland [11], Italy
[32] and Spain [41]; exceptions appeared from the
former Yugoslavia [28], and a couple of papers from
the US, even if these included the only one negative
[24] and a pilot study [22]. On the contrary, these
new data (apart from Italy [29,30]) include the whole
world: Asia (Japan [38,42] and China [31]), Eastern
Europe [33] and the US (but again a pilot
study) [39].
Another aspect we considered, but which must be
looked at carefully, is the result versus patient age
and the duration of treatment (Figure 2). When
applied at the start of puberty – when it is very well
known that the rapidity of progression is greater
than later in the patient’s growth – PEs seem able to
halt or at least reduce the negative evolution. In
older ages, even with very significant curvatures, the
results are positive. Some papers show dramatic
reductions made within very short periods of time
but should be regarded with caution because they
are mainly pilot studies. Another point to consider
carefully is that Figure 2 reports the average results
in the average patients in the average time of
treatment: Variability is not taken in account, while
it is better illustrated in Figure 1. Moreover, the
papers with the largest numbers of patients have
results of stability of the curves (+58) (Figure 3).
Again, this seems to point to the attention given the
quality of the studies, because the size of the sample
is a key factor.
In this review it was possible to divide the exercises
according to the different application (methods)
proposed by the authors. Sub-grouping in systematic
reviews should always be made with caution, even if
it is a big temptation for clinicians. It can be stated
that the only negative paper is clinically the most
stable in many respects; the intrinsic autocorrection
samples generally looked at younger patients with
lower degrees of curvature than the extrinsic auto-
correction ones. Moreover, asymmetric exercises did
show positive results, and a real conclusion regarding
the possible different methods is very difficult to
reach with the actual state of research.
Conclusions
The clinical conclusion is that exercises can be
recommended according to level-1b evidence with
the aim of reducing scoliosis progression. At the
actual stage it is not possible to state anything
regarding the types of exercises to be proposed nor in
regard to the kind of autocorrection to be performed.
A patient of younger age could expect an average
Exercises reduce the progression rate of AIS: Review of literature 783
Disabil Rehabil Downloaded from informahealthcare.com by Dr Stefano Negrini on 03/30/11
For personal use only.
degree of stability even if a great variability should be
accepted, as well as in the latter stage of puberty and
the first stages of adulthood, when on average
reductions can presumably be achieved.
The research conclusion is that solid data coming
from RCTs and long-term observational studies will
be required. Moreover, it is necessary to make a
comparison between different techniques or, even
better, to use a common language in overcoming the
concept of ‘techniques’ and arrive at that of ‘aims of
treatment’ and ‘kind of exercises.’ No data exists
regarding exercises when braces are used (only one
paper discusses exercises as a means to prepare for
bracing), and again this is important. Finally, for the
future we strongly suggest the use of a primary
outcome such as ‘brace prescription’ together with
the secondary outcomes ‘Cobb degrees’ and ‘pro-
gression.’ Along the same line, the evaluation of
QOL (Quality of Life) during treatment should be
done [46,48,49].
References
1. Roach JW. Adolescent idiopathic scoliosis: Nonsurgical
treatment. In: Weinstein SL, editor. The pediatric spine:
Principles and practice. New York: Raven Press; 1997.
pp 497 – 510.
2. Adams F. Hippocrates. The genuine works of Hippocrates.
Translated by Francis Adams, editor. Baltimore: The Wil-
liams and WIlkins Company; 1939.
3. Moen KY, Nachemson AL. Treatment of scoliosis. An
historical perspective. Spine 1999;24:2570 – 2575.
4. Pirola V. La chinesiterapia nella rieducazione della scoliosied.
Milano: Sperling & Kupfer; 1999.
5. Lehnert-Schroth C. Three-dimensional treatment for scolio-
sis. The Schroth Orthopedic Breathing System. A physiother-
apeutic method to improve deformities of the spine. Germany:
Books on Demand; 2007.
6. Stagnara P, Mollon G, De Mauroy J. Reeducation des
scoliosesed. Paris: Expansion Scientifique Francaise; 1990.
7. Mollon G. Kinesite´rapie des scolioses. Encycl Med Chir
26300.A.10 4.6.07.
8. Souchard PE. Posture Me´zie`resed: Mediterraneo Edizioni;
1982.
9. Souchard PE, Ollier M. Le scoliosi. Trattamento fisioterapico
e ortopedico, fisioterapia, tecniche e metodicheed. Torino:
UTET; 2002.
10. Dobosiewicz K, Durmala J, Czernicki K, et al. Pathomechanic
basics of conservative treatment of progressive idiopathic
scoliosis according to Dobosiewicz method based upon
radiologic evaluation. Stud Health Technol Inform 2002;
91:336 – 341.
11. den Boer WA, Anderson PG, v Limbeek J, et al. Treatment of
idiopathic scoliosis with side-shift therapy: An initial compar-
ison with a brace treatment historical cohort. Eur Spine J
1999;8:406 – 410.
12. Negrini S. The evidence-based ISICO approach to spinal
deformities. 1st ed. Milan, Boston: ISICO; 2007.
13. Negrini A, Verzini N, Parzini S, et al. Role of physical exercise
in the treatment of mild idiopathic adolescent scoliosis. Eur
Med Phys 2001:181 – 190.
14. Negrini S, Antonini G, Carabalona R, et al. Physical exercises
as a treatment for adolescent idiopathic scoliosis. A systematic
review. Pediatr Rehabil 2003;6:227 – 235.
15. Lenssinck ML, Frijlink AC, Berger MY, et al. Effect of
bracing and other conservative interventions in the treatment
of idiopathic scoliosis in adolescents: A systematic review of
clinical trials. Phys Ther 2005;85:1329 – 1339.
16. Negrini S, Aulisa L, Ferraro C, et al. Italian guidelines
on rehabilitation treatment of adolescents with scoliosis
or other spinal deformities. Eura Medicophys 2005;41:
183 – 201.
17. Rowe DE, Bernstein SM, Riddick MF, et al. A meta-analysis
of the efficacy of non-operative treatments for idiopathic
scoliosis. J Bone Joint Surg Am 1997;79:664 – 674.
18. Weiss HR, Negrini S, Hawes MC, et al. Physical exercises
in the treatment of idiopathic scoliosis at risk of brace
treatment – SOSORT consensus paper 2005. Scoliosis
2006;1:6.
19. Negrini S, Grivas TB, Kotwicki T, et al. Why do we treat
adolescent idiopathic scoliosis? What we want to obtain and to
avoid for our patients. SOSORT 2005 Consensus paper.
Scoliosis 2006;1:4.
20. Weiss HR. The effect of an exercise program on vital capacity
and rib mobility in patients with idiopathic scoliosis. Spine
1991;16:88 – 93.
21. Athanasopoulos S, Paxinos T, Tsafantakis E, et al. The effect
of aerobic training in girls with idiopathic scoliosis. Scand J
Med Sci Sports 1999;9:36 – 40.
22. Mooney V, Gulick J, Pozos R. A preliminary report on the
effect of measured strength training in adolescent idiopathic
scoliosis. J Spinal Disord 2000;13:102 – 107.
23. Wong MS, Mak AF, Luk KD, et al. Effectiveness of audio-
biofeedback in postural training for adolescent idiopathic
scoliosis patients. Prosthet Orthot Int 2001;25:60 – 70.
24. Stone B, Beekman C, Hall V, et al. The effect of an exercise
program on change in curve in adolescents with minimal
idiopathic scoliosis. A preliminary study. Phys Ther 1979;
59:759 – 763.
25. Weiss HR, Weiss G, Petermann F. Incidence of curvature
progression in idiopathic scoliosis patients treated with
scoliosis in-patient rehabilitation (SIR): An age- and sex-
matched controlled study. Pediatr Rehabil 2003;6:23 – 30.
26. Mollon G, Rodot J. Scolioses structurales mineures et
kine`sitherapie. Etude statistique compareative des re`sultas.
Kinesithe´rapie scientifique 1986:47 – 56.
27. Duconge`P.Lare`e`ducation de la scoliose. Mythe`oure`alite`?
Re`sonance Europeennes Du Rachis 2002:1229 – 1236.
28. Klisic P, Nikolic Z. Scoliotic attitudes and idiopathic scoliosis.
Proceedings of the International Congress on Prevention of
Scoliosis in Schoolchildren. Milan: Edizioni Pro-Juventute,
1985:91 – 92.
29. Negrini S, Negrini A, Romano M, et al. A controlled
prospective study on the efficacy of SEAS.02 exercises in
preventing progression and bracing in mild idiopathic
scoliosis. Stud Health Technol Inform 2006;123:523 – 526.
30. Negrini S, Negrini A, Romano M, et al. A controlled
prospective study on the efficacy of SEAS.02 exercises in
preparation to bracing for idiopathic scoliosis. Stud Health
Technol Inform 2006;123:519 – 522.
31. Wan L, Wang G-x, Bian R. Exercise therapy in treatment of
essential S-shaped scoliosis: Evaluation of Cobb angle in
breast and lumbar segment through a follow-up of half a year.
Zhongguo Linchuang Kangfu (Chinese J Clin Rehabil)
2005;9:82 – 84.
32. Ferraro C, Masiero S, Venturin A. Effect of exercise therapy
on mild idiopathic scoliosis. Europa Medicophysica
1998:25 – 31.
33. Durmala J, Dobosiewicz K, Kotwicki T, et al. Influence of
asymmetric mobilisation of the trunk on the Cobb angle and
rotation in idiopathic scoliosisin children and adolescents.
Ortop Traumatol Rehab 2003;5:80 – 85.
784 S. Negr ini et al.
Disabil Rehabil Downloaded from informahealthcare.com by Dr Stefano Negrini on 03/30/11
For personal use only.
34. Negrini S, Marchini G. Efficacy of the Symmetric, Patient-
oriented, Rigid, Three-dimensional, active (SPoRT) concept
of bracing for scoliosis: A prospective study of the Sforzesco
versus Lyon brace. Eura Medicophys 2007;43(2):171 – 181;
discussion 183 – 184. Epub 2006 Sep 24.
35. Rigo M, Negrini S, Weiss H, et al. SOSORT consensus paper
on brace action: TLSO biomechanics of correction (investi-
gating the rationale for force vector selection). Scoliosis
2006;1:11.
36. Weiss HR, Lohschmidt K, el-Obeidi N, et al. Preliminary
results and worst-case analysis of in patient scoliosis rehabi-
litation. Pediatr Rehabil 1997;1:35 – 40.
37. Otman S, Kose N, Yakut Y. The efficacy of Schroth’s 3-
dimensional exercise therapy in the treatment of adolescent
idiopathic scoliosis in Turkey. Saudi Med J 2005;26:
1429 – 1435.
38. Mamyama T, Kitagawal T, Takeshita K, et al. Side shift
exercise for idiopathic scoliosis after skeletal maturity. Stud
Health Technol Inform 2002;91:361 – 364.
39. McIntire K, Asher M, Burton D, et al. Trunk rotational
strength training for the management of adolescent idiopathic
scoliosis (AIS). Stud Health Technol Inform 2006;123:
273 – 280.
40. Weiss HR. Influence of an in-patient exercise program on
scoliotic curve. Ital J Orthop Traumatol 1992;18:395 – 406.
41. Rigo M, Quera-Salva G, Puigdevall N. Effect of the exclusive
employment of physiotherapy in patients with idiopathic
scoliosis. Proceedings Book of the 11th International Con-
gress of the World Confederation for Physical Therapy.
London, 1991:1319 – 1321.
42. Maruyama T, Matsushita T, Takeshita K, et al. Side shift
exercises for idiopathic scoliosis after skeletal maturity. J.
Bone Joint Surg (Br) 2003;85B.
43. Negrini S, Romano M. Comment on ‘effect of bracing. . .’ by
Lenssinck et al. Phys Ther 2005;85:1329 – 1339. Phys Ther
2007;87:112; author reply 113.
44. Phillips B, Ball C, Sackett D, et al. Oxford Centre for
Evidence-based Medicine Levels of Evidence [Oxford Centre
for Evidence-based Medicine], 2001. Accessed 31 October
2007 from the website: www.cebm.net/index.aspx?o¼1047
45. Reilly RP, Findley TW. Research in physical medicine and
rehabilitation. IV. Some practical designs in applied research.
Am J Phys Med Rehabil 1989;68:196 – 201.
46. White book on physical and rehabilitation medicine in
Europe. Eura Medicophys 2006;42:292 – 332.
47. Negrini S, Minozzi S, Taricco M, et al. A systematic review of
physical and rehabilitation medicine topics as developed by
the Cochrane Collaboration. Eura Medicophys 2007;43:
381 – 390.
48. Weigl M, Cieza A, Cantista P, et al. Determinants of disability
in chronic musculoskeletal health conditions: A literature
review. Eura Medicophys 2007 Nov 9; [Epub ahead of print].
49. Negrini S. Usefulness of disability to sub-classify chronic low
back pain and the crucial role of rehabilitation. Eura
Medicophys 2006;42(3):173 – 175.
Exercises reduce the progression rate of AIS: Review of literature 785
Disabil Rehabil Downloaded from informahealthcare.com by Dr Stefano Negrini on 03/30/11
For personal use only.
