Journal of Bodywork & Movement Therapies (2014) 18, 34e41

Available online at www.sciencedirect.com

journal homepage: www.elsevier.com/jbmt

EXERCISE PHYSIOLOGY

Antagonist coactivation of trunk stabilizer

muscles during Pilates exercises
Denise Martineli Rossi, PT, MSc a,*, Mary Hellen Morcelli, PT,
MSc, Ph.D a, Nise Ribeiro Marques, PT, MSc, Ph.D a,
Camilla Zamfolini Hallal, PT, MSc, Ph.D a, Mauro Gonçalves, PT,
Ph.D a, Dain P. LaRoche, BExSc, Ph.D b, Marcelo Tavella Navega,
PT, Ph.D c

a
Department of Physical Education, São Paulo State University, Rio Claro, Brazil
b
Department of Kinesiology, University of New Hampshire, Durham, NC, USA
c
Department of Physical Therapy and Occupational Therapy, São Paulo State University, Marı´lia, Brazil

Received 6 September 2012; received in revised form 9 April 2013; accepted 11 April 2013

KEYWORDS Summary The purpose of this study was to compare the antagonist coactivation of the local
Electromyography; and global trunk muscles during mat-based exercises of Skilled Modern Pilates. Twelve women
Stability; performed five exercises and concurrently, surface EMG from internal oblique (OI), multifidus
Pilates method; (MU), rectus abdominis (RA) and iliocostalis lumborum (IL) muscles was recorded bilaterally.
Physiotherapy The percentage of antagonist coactivation between local (OI/MU) and global muscles (RA/IL)
was calculated. Individuals new to the practice of these exercises showed differences in coac-
tivation of the trunk muscles between the exercises and these results were not similar bilat-
erally. Thus, in clinical practice, the therapist should be aware of factors such as
compensation and undesirable rotation movements of the trunk. Moreover, the coactivation
of global muscles was higher bilaterally in all exercises analyzed. This suggests that the exer-
cises of Skilled Modern Pilates only should be performed after appropriate learning and correct
execution of all principles, mainly the Centering Principle.
ª 2013 Elsevier Ltd. All rights reserved.

* Corresponding author. Tel.: þ55 19 3526 4345.

E-mail address: denisemartineli@hotmail.com (D.M. Rossi).

1360-8592/$ - see front matter ª 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.jbmt.2013.04.006
Antagonist coactivation of trunk stabilizer muscles during Pilates exercises 35

Introduction regard, electromyography has proven to be an important

tool to study the function of the neuromuscular system of
The Pilates Method is a kinesiotherapeutic technique the low back by quantifying the coactivation of antagonist
widely used by therapists that was created during the First muscle groups (Granata et al., 2005).
World War by Joseph Pilates. The prescription and compo- The stability of the segments of the spine is provided by
sition of the exercises was based on his experience with the the interaction of three subsystems: active, composed by
rehabilitation of injured people (Loss et al., 2010). This skeletal muscles surrounding the spine; passive, composed
method is based on six principles: concentration, control, of vertebrae and their articular structures; and neural,
flowing, precision, breathing and the centering principle, composed of afferent, efferent, and central neural control
considered a differential technique of the method mechanisms (Panjabi, 1992). In physiological conditions,
(Gladwell et al., 2006). The Centering Principle consists of these subsystems provide spinal stability, principally, with
an isometric contraction of the internal oblique and the recruitment of the trunk muscles, which provide dy-
transverse abdominis muscles, which contributes to namic stability (Panjabi, 2003). Moreover, it is suggested
increased antagonist coactivation of the deep lumbar that the most widely used mechanism for the maintenance
muscles (Marques et al., 2012). of spinal stability is the coactivation of trunk muscles, since
Currently, the Pilates Method has been widely applied this strategy promotes protection of spinal structures dur-
in physiotherapy for flexibility training, stimulation of ing the performance of functional activities. In this regard,
blood circulation, improvement of postural alignment electromyography has proven to be an important tool to
and body awareness (Muscolino and Cipriani, 2004). The study the function of the neuromuscular system of the low
exercises challenge the stability of the trunk and acti- back by quantifying the coactivation of antagonist muscle
vate the deep muscles of the lumbo-pelvic region (mul- groups (Granata et al., 2005).
tifidus, internal oblique and transversus abdominis), and Studies of spinal stability point to structural differenti-
require endurance of trunk muscles (rectus abdominis, ation of local and global trunk muscles based on their direct
iliocostalis lumborum and latissimus dorsi) (Endleman and or indirect attachment to the vertebrae. The local muscles,
Critchley, 2008). Thus, the Pilates Method has been used such as multifidus (MU) and internal oblique (OI), despite
for both prevention and rehabilitation of low back pain having a limited capacity to generate torque, have in-
(Muscolino and Cipriani, 2004; Endleman and Critchley, sertions on the vertebrae, which contribute to control the
2008). movement of each lumbar vertebra. On the other hand, the
Low back pain is a common musculoskeletal symptom in global muscles, such as the rectus abdominis (RA) and the
modern society that results in significant costs to health- iliocostalis lumborum (IL), cross several joints with at-
care systems. It is estimated that 90% of cases of low back tachments to the pelvis and the thorax, have a larger
pain have a nonspecific origin, so this condition has a moment arm, and are suited to the control of trunk orien-
multifactorial and complex etiology (Ebenbichler et al., tation and the resistance of external forces (Bergmark,
2001). Among the possible causes of nonspecific low back 1989; Hodges, 2003).
pain are muscle dysfunctions, changes in motor control and In clinical practice, the largest recruitment of global
inadequate recruitment of trunk muscles, which lead to muscles seems to be associated with an increase in spinal
reduced stability of the segments of the spine and altered load that has the potential to cause injury or worsen pain in
distribution of loads in this region (Ebenbichler et al., patients with low back pain (Arokoski et al., 2004). Ac-
2001). cording to Arokoski et al. (2004) therapists should find more
Recently, Skilled Modern Pilates has been proposed as an cautious exercises to challenge the local stabilizing muscles
adaptation of the traditional Pilates Method which has been of the lumbar spine without placing an excessive load on
widely applied in physiotherapy. This new approach to the the structures of this region. Thus, considering the impor-
Pilates Method uses the philosophy created by Joseph tance of lumbopelvic and segmental spinal stability and the
Pilates, but decreases the range of motion of the exercises relationship to the trunk muscle response, the mat-based
and requires a neutral lumbar-pelvic posture (Latey, 2001, Skilled Modern Pilates exercises can be an alternative to
2002). Neutral lumbar-pelvic posture is more adapted to improve conditioning and strength of trunk muscles without
the physiological curvature of the lumbar spine and also overloading the passive structures of the spine (Arokoski
provides an optimal position to increases the recruitment of et al., 2004).
deep lumbar and abdominal muscles (Sapsford et al., 2001; In addition to focusing on the activation of local mus-
O’Sullivan et al., 2006). cles, the Skilled Modern Pilates exercises are performed in
According to O’Sullivan et al. (2006), in the lumbo-pelvic a neutral position that leads to greater local muscle
upright sitting posture, that is, neutral lordosis of the recruitment than global recruitment (O’Sullivan et al.,
lumbar spine, the local muscles, multifidus and internal 2006). Also, these exercises can be considered functional
obliquus, had 62% and 30% higher activity (respectively) since they require the recruitment and sensoriomotor
compared to an extended thoracolumbar spine posture, control of the trunk muscles while performing limb move-
and 62% and 48% higher activity (respectively) compared to ments (Menacho et al., 2010).
a relaxed thoracolumbar spine with a posteriorly rotated To our knowledge, no study had investigated the dif-
pelvis. Additionally, spinal stability is increased with the ferential activation of local and global trunk muscles dur-
coactivation of the local trunk muscles and this strategy ing these exercises. Therefore, the purpose of this study
promotes protection of spine structures during the perfor- was to compare the antagonist coactivation of the local
mance of functional activities (Arokoski et al., 2004). In this and global muscles during several mat-based Skilled Mod-
ern Pilates exercises to ascertain which exercises stimulate
36 D.M. Rossi et al.

local muscle coactivation most. From the principles of control and neutral spinal alignment were taught and
Pilates, we hypothesized that during the execution of the encouraged during the exercises. The neutral position was
exercises in a neutral posture, the local muscle antagonist defined as the intermediate position between lumbar spine
coactivation would be higher than the global muscle retroversion and anteversion and was maintained and
antagonist coactivation to minimize the disturbance of monitored during the exercises by a barometric system
vertebral segments created by upper and lower limb specially developed for the study. This was a simple device
movements. composed of a bag that was inflated under the back of
the volunteers which required them to maintain a certain
Methods pressure (and thus position), with a tolerance of
10 mmHg. The device was used to ensure quality and
precision in exercise performance by testing and moni-
Subjects toring the position of the low back as a control for the
evaluator and also to provide feedback to the subjects
This cross-sectional study utilized twelve, young women, about the contraction of the abdominal muscles.
recruited from the university who were physically active Prior to execution of the exercises, the subjects were
with no previous experience with Pilates exercise (Table 1). placed in an initial position lying supine with their knees
From a pilot study conducted in our laboratory, sample size flexed at 45 , feet aligned in the sagittal plane with the
was estimated by considering a standardized effect size of ischial tuberosity, the lumbar spine in a neutral position,
1.2 with alpha error of 0.05 to find differences in the and the arms extended alongside the body. From this
activation of local and global muscles bilaterally during the starting position, the five exercises were performed, HU I,
Hundreds (level I) exercise. Subjects were excluded for HU II, OLS I, OLS II and SC I, during which the electromyo-
orthopedic disorders, neurological problems, cardiovascu- graphic signal of the local and global muscles of the trunk
lar disease, or previous surgery of the spine or abdomen. were recorded.
This study was approved by the Local Ethics committee The HU I exercise was performed with the subjects in the
(case number 067/2009), and all volunteers signed an supine position, with the hips and knees flexed, while the
informed consent. feet remained flat on the floor and the upper limbs were
held alongside the body. Subjects were instructed to
Procedures perform low amplitude shoulder flexion of both upper limbs
with the hands held approximately 15 cm above the floor
Data collection was performed during two different ses- (Fig. 1A). The HU II exercise involved the same shoulder
sions. On the first day, weight, height, body mass index, age position as HU I, but simultaneously required the subjects
and physical activity level were recorded. Also, the volun- to hold the dominant hip and knee flexed at 90 (Fig. 1B). In
teers were familiarized with the Skilled Modern Pilates the OLS I exercise the subjects performed only lower-limb
Method principles and they practiced all the exercises movements. Starting from the initial flexed position with
performed during the data collection day using the same feet on the ground, the subject performed controlled hip
protocol. On the second day, the subjects performed a and knee extensions, lowering the leg until contact with
single trial of each of the five exercises including, Hundreds the ground, one leg at a time (Fig. 1C). In the OLS II exer-
level I (HU I), Hundreds level II (HU II), One Leg Stretch level cise, the subjects started from the initial position with both
I (OLS I), One Leg Stretch II (OLS II) and Scissors level I (SC hips and knee flexed at 90 with the feet off the ground,
I), in random order. All exercises were supervised by a and then the subjects unilaterally extended the knee and
physiotherapist with experience in the Pilates Method. hip, while keeping the feet off the ground, returning to the
Concurrently, surface EMG from OI, MU, RA and IL was initial position, which was then repeated with the other leg
recorded bilaterally. (Fig. 1D). In the SC I exercise the subjects started with the
knee and hip flexed at 90 bilaterally with both feet off the
Exercises ground. They then touched the toe to the ground, alter-
nating feet, while keeping the contralateral lower limb
The volunteers were taught to recruit their deep abdominal suspended (Fig. 1D).
muscles using a variety of strategies, including visual im-
agery, verbal cueing and demonstration. Moreover the Electromyography
conscious recruitment of the deep abdominal muscles and
the incorporation of the Pilates principles of breathing The surface EMG signal was collected during the exercises
using an eight-channel telemetry system (Noraxon
Table 1 Means of body anthropometrics, age and physical Scottsdale, Arizona, USA) and was recorded with Myor-
activity level of the subjects (n Z 12). esearch software (Noraxon), with a sampling frequency of
2000 Hz and a total gain of 2000 times (20 times in the
Mean (SD) sensor and 100 times in the equipment). The signal was
Mass (kg) 56.74 (7.7) recorded using Ag/AgCl surface electrodes (Miotec, Porto
Height (m) 1.62 (0.06) Alegre, Brazil), placed in bipolar configurations, with a
Body mass index (kg$m2) 21.57 (2.3) recording area of 1 cm in diameter and interelectrode
Age (years) 20.05(2.5) distance of 2 cm. Prior to electrode placement, the skin
Physical activity level (hours/week) 5.54 (1.91) was carefully shaved and cleaned with alcohol to reduce
resistance (Gonçalves et al., 2012).
Antagonist coactivation of trunk stabilizer muscles during Pilates exercises 37

Figure 1 A: Hundreds level I; B: Hundreds level II; C: One Leg Stretch level I; D: One Leg Stretch level II and E: Scissors level I.

The electrodes were placed bilaterally on the muscles as Comparisons were made between percentage coac-
follows: OI, 2 cm medial and inferior to the superior ante- tivation between two local antagonist muscles (OI/MU) and
rior iliac spine; MU, 2 cm lateral to the midline of the two global antagonist muscles of the trunk (RA/IL) as for
interspinous space L4-L5; RA, 3 cm above of the umbilicus the five exercises using repeated measures analysis of
and 2 cm lateral to the midline; IL, 6 cm lateral of the variance (ANOVA) for the right and left sides of the body
intervertebral space L2-L3. A reference electrode was separately. This resulted in a 2 (local vs. global)  5 (ex-
positioned over the styloid process of the ulna on the right ercises) analysis for each side of the body. Post hoc tests
arm (Hermens et al., 2000; Marshall and Murphy, 2003; were performed using the Least Significant Difference (LSD)
Marques et al., 2012). statistic to show which exercise had higher antagonist
coactivation. Statistical significant was set at p < 0.05 for
Data analysis all tests.

EMG signal analysis was carried out using specific routines Results
developed in Matlab (Mathworks Natick, USA). The EMG
signal was full-wave rectified and smoothed using a 4th Table 2 describes the repeated measures ANOVA results of
order low-pass filter with a cut off frequency of 10 Hz. The the %COCON of the local antagonist muscles (OI/MU) and
linear envelope EMG data from each muscle, of each global antagonist muscles (RA/IL) in the five exercises
participant, was used to calculate the level of coactivation. analyzed, for the right and left sides of the body. On the
For this analysis, the area under the signal amplitude curve right side, there was a significant difference for the coac-
was determined for the OI, MU, RA and IL muscles and the tivation between the exercises (f Z 5.145 and P Z 0.002)
percentage of agonist/antagonist coactivation (% COCON) and between the local and global antagonist coactivation
for the local muscles OI/MU and global muscles RA/IL was (f Z 31.966 and P < 0.0001). On the left side, there was a
calculated using the following equation (Candotti et al., difference only for the %COCON between local (OI/MU) and
2009):

2  common area AB
% COCON Z  100
area A þ area B
Table 2 Repeated measures analysis of variance results
where % COCON is the percent coactivation between two for comparison of local and global antagonist coactivation in
antagonist muscles, area A is the area below the smoothed five exercises analyzed for the right and left side.
EMG curve of muscle A, area B the area below the smoothed
Right Side Left Side
EMG curve of muscle B, common area A & B is the common
area of activity of muscles A and B (Candotti et al., 2009). f P f p
Exercises 5. 145 0. 002* 1. 645 0. 180
Statistical analysis Dof: 4
Local/global 31. 966 0. 0001* 37. 336 0. 0001*
The PASW 18.0 statistical package (SPSS Inc.) was used to Dof: 1
perform the statistical analysis. Comparisons were made Exercises  local/ 0. 198 0. 624 0. 753 0. 753
for percentage coactivation between local and global global Dof: 4
muscles for the five exercises using repeated measures
*p < 0.05.
analysis of variance.
38 D.M. Rossi et al.

Figure 2 (a) Comparison of the local and global %COCON on the right side; (b) Comparison of the local and global %COCON on the
left side. y Z significant difference between global (RA/IL) and local (IO/MU) %COCON.

global (RA/IL) antagonist muscles (f Z 37.366 and P < antagonist coactivation of the trunk muscles, particularly
0.001). For both sides, there was no significant difference with respect to the potential vertebral overload during
in the interaction of exercise and antagonist coactivation. these exercises. Activation of trunk muscles has been
For the right side, global %COCON (RA/IL) was 52.01% described as one of the mechanisms used to maintain spinal
higher than local %COCON (IO/MU) (Fig. 2a) and for the left stability and provide protection to structures. This is
side global %COCON was 45.46% higher than local %COCON particularly the case for the local spinal muscles used to
(Fig. 2b). Additionally, HU I exercise had 19.99% and 22.32% minimize the disturbance of the vertebral segments
higher coactivation than OLS I (p Z 0.057) and SC I created by movements of the upper and lower limbs
(p Z 0.013); and HU II had 19.45%, 15.47% and 21.87% (Granata et al., 2005). From the principles of the Pilates
higher coactivation than OLS I (p Z 0.032); OLS II Method, which aim to activate the deep muscles of the
(p Z 0.058) and SC I (p Z 0.011; Fig. 3). lumbo-pelvic region, the initial hypothesis of this study was
that during the exercises in neutral lumbar spinal align-
ment, the antagonist coactivation of the local muscles
Discussion would be higher compared to global muscles.
In this study, the comparison of antagonist coactivation
Due to the significance of the Pilates Method to the clinical of the trunk muscles during mat-based Skilled Modern
population with low back pain, is important to study the Pilates exercises demonstrated that for the right side, the
Antagonist coactivation of trunk stabilizer muscles during Pilates exercises 39

Figure 3 Comparison of the coactivation between the exercises on the right side. * Z significant difference between HU
I Z Hundreds level I and OLS I Z One Leg Stretch level I; £ Z significant difference between HU I and SC I Z Scissors level I;
x
Z significant difference between HU II Z Hundreds level II and OLS I; U Z significant difference between HU II and OLS II Z One
Leg Stretch level II; z Z significant difference between HU II and SC I Z Scissors level I.

HU I and HU II exercises had higher antagonist coactivation. neutral position of the lumbar spine, the highest global
In contrast to our findings, Souza et al. (2001) found no antagonist coactivation of the trunk muscles was necessary
difference in EMG activity between the right and left sides for the production of flexor torque as a result of dynamic
during the execution of exercises for spinal segmental alternation of leg movements and the long lever arm of the
stabilization. Our results may be related to possible lower limb (Hodges, 2003; McCook et al., 2009).
compensation in the trunk muscle activation to maintain The control of joint and spinal stability depends directly
stability during the execution of the exercises. The asym- on the active system coordinated by the contraction of
metry of the load application from the weight of the lower muscles. The stability of the system is provided by the
limbs during the execution of the unilateral exercises necessary stiffness to appropriately restrict movement. In
generates disturbance and trunk rotation which may have the spine, the compressive forces caused by the tension of
led to higher coactivation of trunk musculature on the right passive structures and muscle action in the joints provide
side. Also, the fact that all the subjects of the sample were stiffness and stability. The stabilization of the lumbar spine
right-handed should also be considered. is significantly increased by the tension applied by the local
When investigating the EMG coactivation of the global and and global muscles in the thoracolumbar fascia (Van dieen
local trunk muscles, in contrast to the initial hypothesis, this et al., 2003a, b; Reeves et al., 2007). According to Van
study found that on both sides of the body, antagonist dieen et al. (2003a,b) the direction of the external pertur-
coactivation of the global muscles, RA and IL, was higher in all bation forces is the main factor that will determine which
exercises. The anterior and posterior muscles of the trunk, trunk muscles are recruited to stabilize the spine. Further-
like RA and IL, play important roles in maintaining the sta- more, despite histological, anatomical, and functional dif-
bility of the vertebral column. Because they are considered ferences, the stability of the column is provided by all the
histological and anatomically as global muscles, they are trunk muscles. Thus, the system continuously adjusts the
functionally responsible for generating the torque required stiffness of the column to generate the stability required of
to move the trunk and limbs and to transfer the external certain tasks, and the force and movement characteristics
loads to minimize overload of the spine (Hodges, 2003). of the task determine which muscles are more active (Van
During the execution of OLS level I and II and SC level I dieen et al., 2003a, b, Reeves et al., 2007).
exercises, alternating movements of the lower limbs in Pilates exercises are used for spinal stabilization and
both closed kinetic chain (OLS I) and open kinetic chain have been widely used to treat low back pain (Arokoski
(OLS II and SC I) exercises were performed. There is a et al., 2004). It has been shown that the maintenance of
relationship between increased antagonist coactivation of postural alignment during the exercises stimulates the
the trunk muscles, the asymmetric load applied, and the activation of stabilizing muscles, such as the internal obli-
length of the lever arm (McCook et al., 2009). The local que, transversus abdominis and multifidus muscles (Colado
muscles, MU and OI, act to “fine tune” intervetebral et al., 2011). In the initial phase of rehabilitation,
movements and generate shear forces during changes in the contemporary approaches involve the recruitment of
position of the trunk. The global muscles, RA and IL, are abdominal muscles such as transversus abdominis (TrA) and
fundamental to pelvic stability during exercises with knee OI, with minimal activity of the superficial abdominal
extension and also work to control the torque generated muscles. These practices are based on evidence that in-
by hip extension (Hodges, 2003). Thus, it can be inferred dividuals with low back pain have dysfunction of these
that besides the need to maintain spinal stability and the muscles and exercises that aim to improve the recruitment
40 D.M. Rossi et al.

of these muscles will contribute to vertebral control. activation response may be different in those familiar with
Furthermore, these treatments focus on low-level the exercises. Additional research should be conducted to
contraction, as it has been suggested that low levels of study the activation of trunk muscles in people who
contraction are sufficient to provide the stiffness required routinely perform Pilates exercises to understand the ef-
for control of the intervertebral joints (Urquhart et al., fect of training on trunk stabilizer muscle activation.
2005). However, despite the known beneficial effects of
these exercises, such as reduced intensity of pain and
dysfunction, there is insufficient scientific and clinical ev-
Conflict of interest statement
idence of how the actions of the muscles stabilizing the
spine change during the execution of the exercises None.
(Arokoski et al., 2004; Critchley et al., 2011).
This study examined a sample that did not report back Acknowledgements
pain, however, there is evidence that exercise therapy
produces lower activity of trunk muscles in subjects with
D.M. Rossi was supported by Coordenação de Aperfeiçoa-
low back pain compared with subjects who never had a
mento de Pessoal de Nı́vel Superior (CAPES).
previous painful episode (Arokoski et al., 2004). In clinical
practice, the increased recruitment of global muscles
seems to be associated with increased spinal load that can References
cause injury or worsen the pain in individuals with low back
pain (Arokoski et al., 2004). Although the recruitment of Arokoski, J.P., Valta, T., Kankaanpa, M., Airaksinen, O., 2004.
local muscles is emphasized in the initial phase of reha- Activation of lumbar paraspinal iand abdominal muscles during
bilitation, all the muscles of the trunk are considered therapeutic exercises in chronic low back pain patients.
important for the restoration of normal function and the Archives of Physical Medicine and Rehabilitation 85, 823e832.
progression of exercises involves strategies for the reha- Bergmark, A., 1989. Stability of the lumbar spine: a study in me-
bilitation of the muscular system of the trunk as a whole chanical engineering. Acta Orthopaedica Scandinavica 60, 4e54.
(Arokoski et al., 2004; Urquhart et al., 2005). Candotti, C.T., Loss, J.F., Bagatini, D., Soares, D.P., Rocha, E.K.,
Oliveira, A.R., Guimarães, A.C.S., 2009. Cocontraction and econ-
omy of triathletes and cyclists at different cadences during cycling
Conclusion motion. Journal of Electromyography and Kinesiology 19, 915e921.
Colado, J.C., Pablos, C., Chulvi-Medrano, I., Garcia-Masso, X.,
Flandez, J., Behm, D.G., 2011. The progression of paraspinal
Individuals new to the practice of the mat-based Skilled
muscle recruitment intensity in localized and global strength
Modern Pilates exercises showed differences in coactivation training exercises is not based on instability alone. Archives of
of the trunk muscles between the exercises. These results Physical Medicine and Rehabilitation 92, 1875e1883.
weren’t similar between the sides of the body, because for Critchley, D.J., Pierson, Z., Battersby, G., 2011. Effect of the
only the right side, the HU I and HU II exercises had higher Pilates mat exercises and conventional exercise programmes on
antagonist coactivation. Therefore, in clinical practice, the transverses abdominis and obliquus internus abdominis activity:
physiotherapist should be aware of factors such as pilot randomized trial. Manual Therapy 16, 183e189.
compensation for weak muscles by one side of the body Ebenbichler, G.R., Oddsson, L.I., Kollmitzer, J., Erim, Z., 2001.
that may result in undesirable rotation movements of the Sensory-motor control of the lower back implications for reha-
trunk. Furthermore, the antagonist coactivation of global bilitation. Medicine and Science in Sports and Exercise 33 (11),
1889e1898.
muscles was higher than local muscles in all of the exercises
Endleman, I., Critchley, D.J., 2008. Transversus abdominis and
and it seems that this is associated with increased spinal obliquus internus activity during Pilates exercises: measure-
load. Thus, these exercises should be performed after ment with ultrasound scanning. Archives of Physical Medicine
appropriate learning of the movements and correct and Rehabilitation 89, 2205e2212.
execution of the Pilates principles, mainly the Centering Gladwell, V., Head, S., Haggar, M., Beneke, R., 2006. Does a pro-
Principle, which is to activate the local stabilizing muscles gram of the Pilates improve chronic non-specific low back pain?
to avoid overloading the spine. Journal Sport Rehabilitation 15, 338e350.
Gonçalves, M., Marques, N.R., Hallal, C.Z., Van Dieen, J.H., 2012.
Electromyographic activity of trunk muscles during exercises
Limitations with flexible and non-flexible poles. Journal of Back and
Musculoskeletal Rehabilitation 24, 209e214.
A limitation of this study is that the protocol was performed Granata, K., Lee, P.E., Franklin, T.C., 2005. Co-contraction
in a single attempt and results may have differed if the recruitment and spinal load during isometric trunk flexion and
exercises were performed on separate days. Additional in- extension. Clinical Biomechanics 20 (10), 1029e1037.
formation of how muscle activation is controlled during the Hermens, J.H., Freriks, B., Disselhorst-Klug, C., Rau, G., 2000.
exercises could have been made if the activity of other Development of recommendations for SEMG sensors and sensor
muscles had been studied, such as the external obliques placement procedures. Journal of Electromyography and Kine-
siology 10, 361e374.
and stabilizers of the hip. LSD post hoc tests were used to
Hodges, P.W., 2003. Core stability exercise in chronic low back
increase the statistical power due a relatively small sample pain. Orthopedic Clinics of North America 34, 245e254.
size. Therefore, these initial findings should be interpreted Latey, P., 2002. Updating the principles of the Pilates Method: part
cautiously and the study should be replicated. Another 2. Journal of Bodywork Movement Therapies 6 (2), 94e101.
study limitation is that the volunteers were novices and had Latey, P., 2001. The Pilates Method: history and philosophy. Journal
no experience with the Pilates Method, and the muscle of Bodywork and Movement Therapies 5 (4), 275e282.
Antagonist coactivation of trunk stabilizer muscles during Pilates exercises 41

Loss, J.F., Melo, M.O., Rosa, C.H., Santos, A.B., La Torre, M., Panjabi, M.M., 1992. The stabilizing of the spine. Part I. Function,
Silva, Y.O., 2010. Eletrical activity of external obliqúe and mul- dysfunction, adaptation and enhancement. Journal of Spinal
tifidus muscles during hip flexion-extension exercise performed in Disorders 5 (4), 383e389.
the Cadillac with different adjustments of springs and individual Panjabi, M.M., 2003. Clinial spinal instability and low back
positions. Revista Brasileira de Fisioterapia 14 (6), 510e517. pain. Journal of Electromyography and Kinesiology 13,
Marques, N.R., Morcelli, M.H., Hallal, N.R., Gonçalves, M., 2012. 371e379.
EMG activity of trunk stabilizer muscles during Centering Prin- Reeves, N.P., Narendra, K.S., Cholewicki, J., 2007. Spine stability:
ciple of Pilates Method. Journal of Bodywork and Movements six blind men 493 and the elephant. Clinical Biomechanics 22,
Therapies, Impress. 266e274.
Marshall, P., Murphy, B., 2003. The validity and reliability of sur- Sapsford, R.R., Hodges, P.W., Richardson, C.A., Cooper, D.H.,
face EMG to assess the neuromuscular response of the abdom- Markewell, S.J., Jull, G.A., 2001. Co-activation of the abdom-
inal muscles to rapid limb movement. Journal of inal and pelvic floor muscles during voluntary exercises.
Electromyography and Kinesiology 13, 477e489. Neurology & Urodynamics 20 (1), 31e42.
McCook, D.T., Vicenzino, B., Hodges, P.W., 2009. Activity of deep Souza, G.M., Baker, L.L., Powers, C.M., 2001. Eletromyographic
abdominal muscles increases during submaximal flexion and activity of selected trunk muscles during dynamic spine stabi-
extension efforts but antagonist co-contraction remains unchanged. lization exercises. Archives of Physical Medicine and Rehabili-
Journal of Electromyography and Kinesiology 19, 754e762. tation 82, 86e92.
Menacho, M.O., Obara, K., Conceição, J., Chitolina, M.L., Urquhart, D.M., Hodges, P.W., Allen, T.J., Story, I.H., 2005.
Krantz, D.R., Silva, R.A., Cardoso, J.R., 2010. Electromyo- Abdominal muscle recruitment during a range of voluntary ex-
graphic effect of mat Pilates exercise on the muscle activity of ercises. Manual Therapy 10, 144e153.
the healthy adult females. Journal of Manipulative and Physi- Van dieen, J., Kingma, I., Van Der Gub, J.C.E., 2003a. Evidence
ological Therapeutics 33 (9), 672e678. for a role of antagonistic cocontraction in controlling trunk
Muscolino, J.E., Cipriani, S., 2004. Pilates and the powerhouse-II. stiffness during lifting. Journal of Biomechanics 36,
Journal of Bodywork and Movement Therapies 8, 122e130. 1829e1836.
O’Sullivan, P., Dankaerts, W., Burnett, A.F., Farrel, G.T., Van dieen, J.H., Cholewicki, J., Radebold, A., 2003b. Trunk muscle
Jefford, E., Naylor, C.S., 2006. Effect of different upright pos- recruitment patterns in patients with low back enhance the
tures on spinal-pelvic curvature and trunk muscle activation in a stability of lumbar spine. Spine 28, 834e841.
pain free population. Spine 31 (19), E707eE712.