lable at ScienceDirect

Manual Therapy 16 (2011) 578e584

Contents lists avai

Manual Therapy

journal homepage: www.elsevier .com/math

Original article

The effect of lumbar posture on abdominal muscle thickness during an isometricleg task in people with and without non-specific low back painq

Rafael Zambelli Pinto a,e,*, Paulo Henrique Ferreira b, Marcia Rodrigues Franco c, Mariana Calais Ferreira d,Manuela Loureiro Ferreira e, Luci Fuscaldi Teixeira-Salmela a, Vinicius C. Oliveira b, Christopher Maher e

aDepartamento de Fisioterapia, Universidade Federal de Minas Gerais, MG, BrazilbDiscipline of Physiotherapy, Faculty of Health Sciences, University of Sydney, Sydney, AustraliacRegional Public Hospital of Betim, MG, Brazild Spine Fisioterapia, MG, Brazile The George Institute for Global Health, Sydney Medical School, University of Sydney, Sydney, Australia

a r t i c l e i n f o

Article history:Received 21 October 2010Received in revised form16 May 2011Accepted 23 May 2011

Keywords:Low back painAbdominal musclesUltrasonographyLumbar spine

q The study should be attributed to the Departmeversidade Federal de Minas Gerais (UFMG) e BraTherapy, Universidade Federal de Minas Gerais (UFMEscola de Educação Física, Fisioterapia e Terapia OcupHorizonte, MG, Brazil.* Corresponding author. The George Institute for G

Missenden Rd, Camperdown, Sydney, New South Wal9657 0396.

E-mail address: rafaelzambelli@gmail.com (R.Z. Pi

1356-689X/$ e see front matter � 2011 Elsevier Ltd.doi:10.1016/j.math.2011.05.010

a b s t r a c t

This study investigated the effect of lumbar posture on function of transversus abdominis (TrA) andobliquus internus (OI) in people with and without non-specific low back pain (LBP) during a lower limbtask. Rehabilitative ultrasound was used to measure thickness change of TrA and OI during a lower limbtask that challenged the stability of the spine. Measures were taken in supine in neutral and flexedlumbar postures in 30 patients and 30 healthy subjects. Data were analysed using a two-way (groups,postures) ANOVA. Our results showed that lumbar posture influenced percent thickness change of theTRA muscle but not for OI. An interaction between group and posture was found for TrA thickness change(F1,56 ¼ 6.818, p ¼ 0.012). For this muscle, only healthy participants showed greater thickness changewith neutral posture compared to flexed (mean difference ¼ 6.2%; 95% CI: 3.1e9.3%; p < 0.001).Comparisons between groups for both muscles were not significant. Neutral lumbar posture can facilitatean increase in thickness of the TrA muscle while performing a leg task, however this effect was notobserved for this muscle in patients with LBP. No significant difference in TrA and OI thickness changebetween people with and without non-specific LBP was found.

� 2011 Elsevier Ltd. All rights reserved.

1. Introduction

Chronic non-specific low back pain (LBP) can be defined as painand disability persisting for more than 3 months without a clearspecific cause (van Tulder et al., 2002). The condition is highlyprevalent in industrialized societies and carries a significanteconomic and social burden (Waddell, 2004). A great variety ofinterventions has been proposed for the treatment of LBP, includingactive treatments such as general exercise (Moffett et al., 1999) andmotor control exercise (Hodges, 2003); electrotherapy (Khadilkar

nt of Physical Therapy, Uni-zil. Department of PhysicalG), Av. Antônio Carlos 6627,acional, CEP: 31270-010, Belo

lobal Health, PO Box M201,es, 2050 Australia. Tel.: þ61 2

nto).

All rights reserved.

et al., 2008); spinal manipulative therapy (Bronfort et al., 2004);advice and education (Heymans et al., 2005); and cognitive-behavioural therapy (Henschke et al., 2010). Among all theseinterventions, motor control training or exercise has becomeincreasingly popular and there is evidence of effectiveness for thetreatment of chronic LBP (Ferreira et al., 2006, 2007; Costa et al.,2009a; Macedo et al., 2009).

Motor control training aims to restore the coordination andcontrol of the spine. The treatment rationale is that the pattern oftrunk muscle recruitment and postural control may be altered inpatients with LBP (Richardson and Jull, 1995; O’Sullivan, 2000;Hodges, 2003). There is some evidence of altered abdominalmuscle recruitment in people with LBP affecting specific deepstabilizing muscles such as transversus abdominis (TrA) (Hodgesand Richardson, 1996, 1998, 1999), obliquus internus (OI) (Hodgesand Richardson, 1996, 1998, 1999), multifidus (Hides et al., 1994),diaphragm and pelvic floor muscles (O’Sullivan et al., 2002a). Thealtered recruitment pattern of TrA and OI in people with LBP ischaracterized by delayed onset (Hodges and Richardson, 1996,

R.Z. Pinto et al. / Manual Therapy 16 (2011) 578e584 579

1998, 1999) and reduced thickness change (Ferreira et al., 2004;Teyhen et al., 2009b) during voluntary tasks involving the upper(Hodges and Richardson, 1996; Hodges and Richardson, 1999) andlower limbs (Hodges and Richardson, 1998; Ferreira et al., 2004;Teyhen et al., 2009b). Recently, studies have shown that interven-tions targeting specific retraining of the deep abdominal musclesare effective in restoring the timing of TrA onset (Tsao and Hodges,2008) and activation of TrA (Ferreira et al., 2010). Furthermore, theimprovement in TrA thickness change correlates with the decreasein disability in people with LBP (Ferreira et al., 2010).

A variety of clinical strategies are used by physical therapists toretrain the deep abdominal muscles, including; indirect palpationof TrA (Costa et al., 2006), verbal cueing (Critchley, 2002) and visualinformation by ultrasound imaging biofeedback (Henry andWestervelt, 2005; Costa et al., 2009b). Another type of strategycommonly used in the retraining of the deep spinal muscles is toemphasize a neutral lumbar posture while performing motorcontrol training (O’Sullivan, 2000). The clinical definition of neutrallumbar posture involves neutral pelvic tilt, while maintaining thelordosis of lumbar and cervical spines with a smooth transition intothoracic kyphosis (Hodges et al., 2009). Moreover, it has beenadvocated that upright sitting favours the control of the neutralzone by requiring the recruitment of deep abdominal muscles and,consequently, unloading the spinal column components (disk,ligaments and facets) (Panjabi, 2003).

The upright sitting posture has been shown to promote anincrease in electromyographic (EMG) activity of cervical and lum-bopelvic muscles with a postural stabilizing role, such as multifidus(O’Sullivan et al., 2002b, 2006; Falla et al., 2007; Claus et al., 2009),deep cervical flexors (Falla et al., 2007) and pelvic floor muscles(Sapsford et al., 2006). Moreover, a controlled randomized clinicaltrial has shown that an intervention approach which emphasizesthe control of the lumbar neutral zone in different tasks is capableof reducing pain in people with LBP (Suni et al., 2006). In contrastthe flexed lumbar posture inhibits the activation of multifidus inhealthy subjects (O’Sullivan et al., 2002b, 2006) and it has beenclinically associated with exacerbation of pain in people with LBP(O’Sullivan, 2000).

To the best of our knowledge, however, no studies have inves-tigated the effect of neutral and flexed lumbar postures on TrA andOI function during voluntary leg tasks. This is an importantknowledge gap because many motor control training programsinclude voluntary leg tasks as a challenge to the control of the spineand pelvis. Therefore, the aim of this study was to investigate theeffect of two different lumbar spine postures, neutral and flexedlumbar posture, on the thickness change of the TrA and OI in peoplewith or without LBP during a leg task.

2. Methods

2.1. Participants

Sixty study participants (30 with chronic LBP and 30 healthysubjects) volunteered for this study. Participants were excludedfrom the control group if they had a history of low back pain thathad restricted normal daily activities or caused them to have timeoff work, if they had any respiratory or neurologic disorder or painelsewhere in the spine or lower limbs. Participants were includedin the LBP group if they had chronic non-specific low back pain,defined as pain lasting for at least 3 months without clear specificcause, with or without pain referral to the leg, but withoutneurological deficit. Exclusion criteria for LBP group were: spinaland abdominal surgery in the past 12 months; a report of preg-nancy; suspected or diagnosed serious spine pathology (inflam-matory spondyloarthropathies, fracture, malignancy, cauda

equina syndrome, or infection) or nerve root compromise (diag-nosed by at least two positive tests out of the following: reflextests, sensation tests, and muscle power tests). For descriptivepurposes disability, pain, kinesiophobia and function wereassessed using the Brazilian-Portuguese versions of the RolandMorris Disability Questionnaire (Nusbaum et al., 2001), the 0e10pain scale (Ross, 1997), the Tampa Scale (Siqueira et al., 2007) andPatient Specific Functional Scale (PSFS) (Westaway et al., 1998),respectively. Ethical approval from the Federal University of MinasGerais e Brazil Ethics Committee and written informed consentwere obtained.

2.2. Procedure and data collection

Abdominal muscle function was measured indirectly usingrehabilitative ultrasound to measure muscle thickness changesaccording to a protocol developed by Ferreira et al. (2004). In thepresent study, this protocol had to be adapted to the two experi-mental conditions: neutral lumbar posture and flexed lumbarposture. These two lumbar postures simulate the lumbar curvaturecommonly found in upright and slumped sitting, respectively.Participants were positioned in these two experimental conditionsby placing wedges under the pelvis, in supine lying (Fig. 1). Thewedges were individually selected according to lumbar lordosisangles formed with the subjects in neutral and flexed sittingmeasured by a flexible ruler. Details of the wedge selection processhave been previously published (Pinto et al., 2011).

Before ultrasonographic recordings, measurements of lumbarlordosis angle in both sitting conditions were conducted in order toselect the respectivewedges. Neutral sitting, frequently encouragedby physiotherapists during postural re-education and motorcontrol training, was defined clinically by positioning the partici-pant seated on the ischial tuberosities with the manubrium inalignment with the anterior aspect of the pelvis, involving themaintenance of a neutral pelvic tilt, lumbar and cervical lordosisand smooth transition to thoracic kyphosis under the supervisionof an experienced physiotherapist (R.Z.P.), whereas flexed sittingwas defined as a commonly adopted relaxed posture while sittingin an unsupported chair.

Dimensions of the wedges were determined in a previous pilotstudywith 22 subjects (12 healthy and 10 LBP subjects). In this pilotstudy, all subjects had their lumbar lordosis measured while sittingunsupported on an adjustable stool with their hips at 50�, their feetpositioned shoulder width apart and the arms crossed over thechest in the two sitting conditions. The lumbar curvature angle wasdetermined using the formula described by (Hart and Rose, 1986).Testeretest reliability on different days (one week apart) showedan intraclass correlation coefficient (ICC3,1) of 0.86 with 95%confidence interval (CI) of 0.72e0.94 and standard error of themeasurement (SEM) of 5.93� for the measurement of the lumbarcurvature using a flexible ruler with 40 cm in length (TRIDENT Ltda,Itapuí, SP, Brazil). Positive angles refer to a lordotic curve andnegative angles to a kyphotic curve. The 22 subjects had theirlumbar curvature angles measured in upright and slumped sitting.A total of 44 measures (mean ¼ 25.5�; SD ¼ 19,8�;range ¼ �25�e63�) were divided in five equal parts using thefollowing percentiles: P0 ¼ �25�, P20 ¼ 10�, P40 ¼ 20�, P60 ¼ 32�,P80 ¼ 45� and P100 ¼ 63�. In this manner, the angulations of thefive wedges were manufactured according to the average valuebetween P0eP20, P20eP40, P40eP60, P60eP80, P80eP100, withthe five wedges having the respective angulations e 8�, 15�, 26�,39� and 54�. The four wedges with positive value angles were halfoval-shaped (Fig. 1A) with the following dimensions: length stan-dardized to 14 cm, width standardized to 40 cm and the heights of0.5 cm, 0.8 cm,1.2 cm and 1.7 cm corresponding to the angles of 15�,

Fig. 1. Experimental setup adapted in order to position the lumbar curvature in a neutral lumbar posture and flexed lumbar posture. (A) shows the position of the half oval-shapedwedgewith itsdimension.H represents theheights (0.5 cm, 0.8 cm,1.2 cmor1.7cm) for the fourhalf oval-shapedwedges. (B) showshowthe triangle-shapedwedgewasplaced and itsdimension.

R.Z. Pinto et al. / Manual Therapy 16 (2011) 578e584580

Table 1Wedge selection process.

Interval percentile 0e20 20e40 40e60 60e80 80e100Interval angulation �25e10 10e20� 20�e32� 32�e45� 45�e63�

Wedge angulationa �8� 15� 26� 39� 54�

Wedge height e 0.5 cm 0.8 cm 1.2 cm 1.7 cmWedge shape Triangular-shaped Oval-shaped Oval-shaped Oval-shaped Oval-shaped

Positive angles refer to lordosis and negative angles to kyphosis.a Angulations calculated by the average between the lowest and highest values of the intervals.

Table 2Characteristics of study participants.

Characteristics Healthy subjects Patients with LBP

Age (yr) 41.3 (10.9) 42.9 (13.4)Weight (kg) 70.1 (11.7) 68.8 (11.5)Height (cm) 168.3 (8.7) 167.2 (9.1)BMI (kg/m2) 24.6 (2.7) 24.7 (3.2)Pain (0e10 scale) e 5.5 (1.93)RMDQ (0e24) e 7.3 (3.96)Tampa (17e68) e 36.3 (5.72)PSFS (3e30) e 12.1 (5.35)GenderMales (%)* 17 (57) 10 (33)

Abbreviations: BMI, body mass index; LBP, low back pain; PSFS, Patient SpecificFunctional Scale; RMDQ, Roland Morris Disability Questionnaire.Values are means (SD) unless otherwise denoted.*p < 0.05 for between-group difference.

R.Z. Pinto et al. / Manual Therapy 16 (2011) 578e584 581

26�, 39� and 54�, respectively. A 0.8 mm layer of Ethyl Vinyl Acetatewas used to cover the wedges for comfort purposes and tocompensate for the ruler’s thickness. The only wedge with a nega-tive value would have been fabricated as an inverted wedge but itwas not feasible to lay the subject on an inverted half oval-shapedwedge so we opted for manufacturing a triangle-shaped wedge(Fig. 1B) to be placed at the thoracic spine achieving the desirableflexed posture of the lumbar curvature.

Each participant’s lumbar curvature was measured in (i) neutraland (ii) slumped sitting and then allocated the two wedges closestto the two values to simulate in the supine position the lumbarcurvatures in both sitting postures (Table 1). The allocated wedgeswere then used to control posture in subsequent tests.

After the wedge selection process, ultrasonographic recordingsfor each subject were taken in both conditions, neutral and flexedposture, in a random order following a protocol published previ-ously (Ferreira et al., 2004). Ultrasound images were made witha 10 cm, 7.5 MHz linear-array transducer (Sonoline SL1, SiemensLtda, São Paulo, SP, Brazil). According to this protocol, studyparticipants were positioned in supine on a plinth with armscrossed over the chest, the hips flexed to 50� and knees flexed at90�. In this position, study participants were instructed to performsix lower limb tasks involving two different directions of legmovement, three isometric knee flexion and three isometric kneeextension efforts, to target forces based on 7.5% of their bodyweight measured by a force transducer attached to a belt strappedaround the ankles and a metallic frame positioned at the end of thebed (Fig. 1). The order of direction of leg movement was random-ized and subjects were providedwith verbal feedback of force givenby one of the examiners. The transducer was placed unilaterally onthe right side and transversely across the abdominal wall alonga line midway between the inferior angle of the rib cage and theiliac crest. For each leg task, static ultrasound images were collectedconsecutively at rest (rest image) and once the target isometricknee flexion or extension force (contraction image) had beenreached. A total of 6 pairs of rest/contraction images, 3 pairs foreach direction of leg movement, were collected for each lumbarposture condition (neutral and flexed) and stored for later analysisby a blinded examiner. An intra-tester reliability pilot study with 10healthy subjects and 10 LBP patients for the measurement of TrAand OI thickness change on different days (oneweek apart) showedan ICC3,6 of 0.89 (95% CI: 0.76e0.96; SEM: 2.34%) and 0.75 (95% CI:0.50e0.89; SEM: 1.46%), respectively.

2.3. Data analysis

Ultrasound images were measured with custom designed soft-ware: LabVIEW (National Instruments Corporation, Austin, TX,USA). A grid was placed over the image andmeasures of TrAmusclethickness changewere made at three sites: center of the image, and1 cm (calibrated to the image scale) left and right from the midline.The average of the three measures in each image was recorded foranalysis. The measurement protocol is described in detail else-where (Ferreira et al., 2004). Resting thickness, contracted

thickness and thickness change were used for analysis. Thesemeasures were collected from the six pairs of rest/contractionimages being three pairs for each direction of leg movement.Resting thickness and contracted thickness were defined as theaverage thickness of the abdominal muscles in the 6 rest imagesand 6 contraction images (3 images for each direction of legmovement), respectively. The change in thickness in each pair ofrest/contraction images was expressed as a proportion of thethickness at rest. A final measure of muscle thickness change foreach lumbar posture condition was established by calculating theaverage of the change in thickness considering all six pairs ofimages, three from each direction of leg movement.

2.4. Statistical analysis

To assess for between-group differences, c2 was used for genderand independent t-tests for age, height, weight and body massheight (BMI). Paired t-tests were used to assess the difference inlumbar curvature between upright and slumped sitting within eachgroup.

Ultrasound data, thickness change and resting thickness, wereanalysed using a 2 (groups) � 2 (postures) analysis of variance(ANOVA) with the between subjects factor being group (LBP vs.healthy subjects) and within-subject factor being posture (neutralvs. flexed) for each muscle (i.e. three separate ANOVA). Gender andBMI were included as covariates in the analysis of thickness change.Post hoc testing was undertaken with Fisher’s multiple range testwhen significant interaction effects were identified.

Effects of each factor were expressed as mean (95% CI) differ-ences. Statistical analyses were performed using STATISTICA 7.0software (StatStoft Inc, Tulsa, OK, USA) and alpha level was set at0.05.

3. Results

Table 2 shows participants’ characteristics such as age, height,weight and gender; and LBP group characteristics of disability, pain,

Table 3Lumbar curvature angles in upright and flexed sitting for healthy subjects andpatients with LBP.

Lumbar curvaturein upright sitting(�)

Lumbar curvaturein flexed sitting (�)

Healthy subjects* 32.0 (11.1) 9.2 (16.1)Patients with LBP* 38.6 (10.2) 16.3 (13.9)

Values are means (SD).*P < 0.05 for difference within group.

R.Z. Pinto et al. / Manual Therapy 16 (2011) 578e584582

kinesiophobia and function. A significant difference betweengroups in gender was observed (p < 0.05).

Fig. 2. Muscle thickness change in the two lumbar postures. Thickness change isrepresented by muscle thickness increase expressed as a percentage of baseline. Leftpanel is for healthy subjects and right panel for patients with LBP. Data are means and95% CI for TrA (:) and OI (-). *P < 0.001.

3.1. Lumbar curvature in different sitting postures within groups

The experimental conditions used in this study followeda standardised protocol inwhich wedges were used to simulate thelumbar curvatures found in upright and slumped sitting. Bycomparing the lumbar curvature in both sitting postures mayindicate if patients performed the leg task in two different lumbarpostures. Our findings showed significant differences between bothsitting conditions within each group. The mean difference betweenupright and slumped sitting was 22.8� (95% CI: 26.9e18.7;p < 0.001) for healthy subjects and 22.0� (95% CI: 26.3e17.7;p < 0.001) for people with LBP. Table 3 shows the mean and stan-dard deviations for each sitting condition and group.

3.2. Effect of lumbar posture on abdominal muscle resting thicknessand thickness change in people with and without LBP

The resting thickness, contracted thickness and thicknesschange (expressed as a percentage of baseline) for all muscles ineach lumbar posture (neutral and flexed) and group are shown inTable 4.

Our analyses revealed that resting thickness of the abdominalmuscles were similar across groups and lumbar posture conditions.ANOVA results for resting thickness of all abdominal musclesshowed no significant difference for main effect of group (TrA:F1,58 ¼ 0.366, p ¼ 0.548; OI: F1,58 ¼ 1.429, p ¼ 0.237) and posture(TrA: F1,58 ¼ 1.242, p ¼ 0.230; OI: F1,58 ¼ 2.953, p ¼ 0.091) orinteraction effect of group by posture (TrA: F1,58 ¼ 0.574, p ¼ 0.452;OI: F1,58 ¼ 0.006, p ¼ 0.938).

The effect of lumbar posture on TrA thickness change whileperforming the low load leg task showed different results inhealthy subjects compared to people with LBP. ANOVA resultsshowed a significant group by posture interaction effect(F1,56 ¼ 6.818, p ¼ 0.012). The covariates, gender (F1,56 ¼ 3.154,

Table 4Abdominal muscles resting thickness, contracted thickness and thickness change inneutral and flexed postures for healthy subjects and patients with LBP.

Healthy Subjects Patients with LBP

Neutral Flexed Neutral Flexed

Resting thickness (mm)TrA 3.8 (0.9) 3.9 (0.8) 3.7 (0.9) 3.7 (0.9)OI 8.7 (2.2) 8.8 (2.4) 8.0 (2.2) 8.1 (2.3)Contracted thickness (mm)TrA 4.2 (1.0) 4.1 (0.8) 4.0 (1.0) 3.9 (1.0)OI 9.0 (2.4) 9.0 (2.4) 8.2 (2.2) 8.2 (2.2)Thickness change (%)TrA 10. 3 (9.8) 4.1 (5.8) 6.8 (8.0) 5.9 (7.8)OI 3.4 (3.7) 1.7 (3.7) 2.8 (3.6) 1.0 (2.8)

Values are means (SD).

p ¼ 0.081) and BMI (F1,56 ¼ 0.000, p ¼ 0.991), were not significant.In healthy participants, the effect of neutral lumbar posture on TrAthickness change was significantly greater than flexed (meandifference ¼ 6.2%; 95% CI: 3.1e9.3%; p < 0.001), whereas nosignificant difference in TrA thickness change was found in patientswith LBP (mean difference ¼ 1.0%; 95% CI: �2.1e4.0%; p ¼ 0.537).Unexpectedly, comparisons between healthy subjects and peoplewith LBP regarding the automatic increase in TrA thickness changewhile performing the low load leg task in neutral (meandifference ¼ 3.4%; 95% CI: �1.5e8.3%; p ¼ 0.169) or in flexedposture (mean difference ¼ �1.8%; 95% CI: �6.7 to 3.1%; p ¼ 0.458)did not show differences between groups (Fig. 2).

The automatic increase in thickness change for OImuscle did notshow any difference between groups or postures. ANOVA results forOI revealed no main effect of group (F1,56 ¼ 1.603, p ¼ 0.211) orposture (F1,56 ¼ 0.686, p ¼ 0.411). In contrast with TrA results, nogroup by posture interaction effect (F1,56 ¼ 0.055, p ¼ 0.815) wasfound. The covariates, gender (F1,56 ¼ 2.932, p ¼ 0.092) and BMI(F1,56 ¼ 1.335, p ¼ 0.253), were also not significant.

4. Discussion

According to our findings, an increase in automatic TrA thick-ness during a low load leg task was greater in the neutral lumbarposture compared to the flexed lumbar posture in healthy subjectsbut not in people with LBP. For OI muscle, however, performing thetask in the neutral or the flexed lumbar posture showed similarincrease in automatic thickness change of thismuscle. In contrast toprevious studies, no significant difference in TrA thickness changewas found between people with and without LBP. These resultshighlight the role of the neutral lumbar posture in the recruitmentof the deep abdominal muscles and, surprisingly, challenge theview that patients with LBP have deficits in TrA thickness change.

An automatic increase in thickness change of deep abdominalmuscles measured by rehabilitative ultrasound imaging has beendocumented in healthy subjects while performing leg tasks inunloaded (Ferreira et al., 2004; Hides et al., 2007; Teyhen et al.,2009a) and loaded postures (Ainscough-Potts et al., 2006). Thisincrease in muscle thickness change may be required to counter-balance the destabilizing force generated in the lumbopelviccomplex by the leg task. In the present study, we found that neutrallumbar posture facilitates the increase in automatic TrA thicknesschange during the leg task in healthy subjects compared to flexedlumbar posture. It has been advocated that there is a linear

R.Z. Pinto et al. / Manual Therapy 16 (2011) 578e584 583

correlation between thickness change and electromyographyduring low load contractions for TrA (Hodges et al., 2003;McMeeken et al., 2004) and OI muscles (Hodges et al., 2003).More importantly, the experimental protocol used in this study hasshown strong correlation between these measures for TrA and OImuscles (Ferreira et al., 2011). It may be that the neutral lumbarposture has a role to play in the stabilizing mechanism of thelumbar spine. As spine stability in a flexed posture relies on passiverestraints such as the disk, ligaments and facets (Panjabi, 2003), ithas been suggested that in the neutral posture greater recruitmentof deep stabilizing muscles, such as TrA muscle, may be needed tomeet the demand of optimal stability. This might be true for healthysubjects, however, that strategy was not observed in people withLBP who participated in this study. In this case, symptomaticpatients may use different postural adjustments, such as co-contraction of superficial muscles (van Dieen et al., 2003), toenhance spinal stability.

An unexpected finding of the study was that there was nodifference in automatic TrA thickness change between people withand without LBP. This result would seem to challenge a previousstudy (Ferreira et al., 2004) which has reported deficits in TrAthickness in people with LBP measured by the same protocol. Inagreement with our findings, another study (Hides et al., 2009)found no difference between groups in automatic TrA thicknesschange during another leg task, unilateral weight-bearing task, ina similar unloaded posture. It is possible that the low load task usedin these studies were not able to demonstrate difference in auto-matic TrA thickness change between both groups. When investi-gating a different task, the voluntary contraction, a previous studyfrom our group using the same sample of participants showed alsono difference in TrA thickness change between groups (Pinto et al.,2011). We suggest future studies compare specific sub-groups ofpatients, as, for instance, patients with unilateral LBP and a positivestraight-leg raise test. Patients in this sub-group have shown lessincrease in TrA thickness change while performing a voluntarymanoeuver (Teyhen et al., 2009a) or actively raising their leg(Teyhen et al., 2009b). Our results are consistent with the emergingview that not all patients with LBP have a reduced ability to recruitthese stabilizing muscles. Therefore, at present, further studies areneeded to shed light on whether this impairment would exist ina specific sub-group of patients or would be elicited by higher loadtasks. Logically interventions to improve recruitment of theabdominal stabilizing muscles would seem to only be indicated inpatients clearly with this specific impairment. This highlights theimportance of sensitive clinical tools to identify patients with theseimpairments who may be candidates for motor control exercise.

In the current study, we developed a standardized protocol tosimulate in supine lying the lumbar curvature angles found in twomeaningful clinical sitting postures. The significant differencefound between upright and slumped sitting within each group(Table 3) gives support to the notion that we measured abdominalmuscle thickness in different lumbar postures. One problem is thatby altering the lumbar posture, changes the intra-abdominalpressure might occur, which is likely to affect the resting thick-ness of the muscles. However, our analysis might not have beeninfluenced by these factors as no difference in resting thickness ofabdominal muscles across conditions and groups was observed.Furthermore, gender and BMI are also likely to influence the restingthickness of a muscle (Springer et al., 2006). Although we useda non-matched group design on gender and BMI, all analysisrevealed no influence of these factors. Nevertheless, our testingprotocol is considered to be a non-functional posture. Moredefinitive answers regarding the influence of lumbar posture ondeep muscle function might also need to consider more functionalpostures and limb tasks.

5. Conclusion

Our study demonstrates that the neutral lumbar posture canfacilitate an increase in thickness of the TrA muscle while per-forming a leg task, however this effect was not observed for thismuscle in patients with LBP. Thus, in healthy subjects the neutralposture may be needed to meet the demand of optimal stability. Incontrast, OI muscle function during the leg task was not influencedby lumbar posture. Another finding of this study was that nodifference in the lateral abdominal muscle thickness changebetween people with and without LBP was found. Further studiesfocussing on clinical strategies and sub-groups of patients withspecific impairments may contribute to a greater understanding of,and perhaps clinical role for, motor control training.

References

Ainscough-Potts AM, Morrissey MC, Critchley D. The response of the transverseabdominis and internal oblique muscles to different postures. Manual Therapy2006;11(1):54e60.

Bronfort G, Haas M, Evans RL, Bouter LM. Efficacy of spinal manipulation andmobilization for low back pain and neck pain: a systematic review and bestevidence synthesis. Spine Journal 2004;4(3):335e56.

Claus AP, Hides JA, Moseley GL, Hodges PW. Different ways to balance the spine:subtle changes in sagittal spinal curves affect regional muscle activity. Spine2009;34(6):E208e14.

Costa LO, Costa Lda C, Cancado RL, Oliveira Wde M, Ferreira PH. Short report: intra-tester reliability of two clinical tests of transversus abdominis muscle recruit-ment. Physiotherapy Research International 2006;11(1):48e50.

Costa LO, Maher CG, Latimer J, Hodges PW, Herbert RD, Refshauge KM, et al. Motorcontrol exercise for chronic low back pain: a randomized placebo-controlledtrial. Physical Therapy 2009a;89(12):1275e86.

Costa LO, Maher CG, Latimer J, Smeets RJ. Reproducibility of rehabilitative ultra-sound imaging for the measurement of abdominal muscle activity: a systematicreview. Physical Therapy 2009b;89(8):756e69.

Critchley D. Instructing pelvic floor contraction facilitates transversus abdoministhickness increase during low-abdominal hollowing. Physiotherapy ResearchInternational 2002;7(2):65e75.

Falla D, O’Leary S, Fagan A, Jull G. Recruitment of the deep cervical flexor musclesduring a postural-correction exercise performed in sitting. Manual Therapy2007;12(2):139e43.

Ferreira PH, Ferreira ML, Hodges PW. Changes in recruitment of the abdominalmuscles in people with low back pain: ultrasound measurement of muscleactivity. Spine 2004;29(22):2560e6.

Ferreira PH, Ferreira ML, Maher CG, Herbert RD, Refshauge K. Specific stabilisationexercise for spinal and pelvic pain: a systematic review. Australian Journal ofPhysiotherapy 2006;52(2):79e88.

Ferreira ML, Ferreira PH, Latimer J, Herbert RD, Hodges PW, Jennings MD, et al.Comparison of general exercise, motor control exercise and spinal manipulativetherapy for chronic low back pain: a randomized trial. Pain 2007;131(1e2):31e7.

Ferreira P, Ferreira M, Maher C, Refshauge K, Herbert R, Hodges P. Changes inrecruitment of transversus abdominis correlate with disability in people withchronic low back pain. British Journal of Sports Medicine 2010;44(16):1166e72.

Ferreira PH, Ferreira ML, Nascimento DP, Pinto RZ, Franco MR, Hodges PW.Discriminative and reliability analyses of ultrasound measurement of abdom-inal muscles recruitment. Manual Therapy, doi:10.1016/j.math.2011.02.010;2011.

Hart DL, Rose SJ. Reliability of a noninvasive method for measuring the lumbarcurve. Journal of Orthopaedic and Sports Physical Therapy 1986;8(4):180e4.

Henry SM, Westervelt KC. The use of real-time ultrasound feedback in teachingabdominal hollowing exercises to healthy subjects. Journal of Orthopaedic andSports Physical Therapy 2005;35(6):338e45.

Henschke N, Ostelo RW, van Tulder MW, et al. Behavioural treatment for chroniclow-back pain. Cochrane Database of Systematic Reviews 2010;7:CD002014.

Heymans MW, van Tulder MW, Esmail R, Bombardier C, Koes BW. Back schools fornonspecific low back pain: a systematic review within the framework of theCochrane Collaboration Back Review Group. Spine 2005;30(19):2153e63.

Hides JA, Stokes MJ, Saide M, Jull GA, Cooper DH. Evidence of lumbar multifidusmuscle wasting ipsilateral to symptoms in patients with acute/subacute lowback pain. Spine 1994;19(2):165e72.

Hides JA, Miokovic T, Belavy DL, Stanton WR, Richardson CA. Ultrasound imagingassessment of abdominal muscle function during drawing-in of the abdominalwall: an intrarater reliability study. Journal of Orthopaedic and Sports PhysicalTherapy 2007;37(8):480e6.

Hides JA, Belavy DL, Cassar L, Williams M, Wilson SJ, Richardson CA. Alteredresponse of the anterolateral abdominal muscles to simulated weight-bearingin subjects with low back pain. European Spine Journal 2009;18(3):410e8.

Hodges PW. Core stability exercise in chronic low back pain. Orthopedic Clinics ofNorth America 2003;34(2):245e54.

R.Z. Pinto et al. / Manual Therapy 16 (2011) 578e584584

Hodges PW, Richardson CA. Inefficient muscular stabilization of the lumbar spineassociated with low back pain. A motor control evaluation of transversusabdominis. Spine 1996;21(22):2640e50.

Hodges PW, Richardson CA. Delayed postural contraction of transversus abdominisin low back pain associated with movement of the lower limb. Journal of SpinalDisorders 1998;11(1):46e56.

Hodges PW, Richardson CA. Altered trunk muscle recruitment in people with lowback pain with upper limb movement at different speeds. Archives of PhysicalMedicine and Rehabilitation 1999;80(9):1005e12.

Hodges PW, Pengel LH, Herbert RD, Gandevia SC. Measurement of musclecontraction with ultrasound imaging. Muscle & Nerve 2003;27(6):682e92.

Hodges P, Ferreira P, Ferreira M. Lumbar spine: treatment of instability and disor-ders of movement control. In: Magee DJ, Zachazewski JE, Quillen WS, editors.Pathology and intervention in musculoskeletal rehabilitation. St. Louis: Saun-ders Elsevier; 2009. p. 389e425.

Khadilkar A, Odebiyi DO, Brosseau L, Wells GA. Transcutaneous electrical nervestimulation (TENS) versus placebo for chronic low-back pain. Cochrane Data-base of Systematic Reviews 2008;4:CD003008.

Macedo LG, Maher CG, Latimer J, McAuley JH. Motor control exercise for persistent,nonspecific lowback pain: a systematic review. Physical Therapy 2009;89(1):9e25.

McMeeken JM, Beith ID, Newham DJ, Milligan P, Critchley DJ. The relationshipbetween EMG and change in thickness of transversus abdominis. ClinicalBiomechanics 2004;19(4):337e42.

Moffett JK, Torgerson D, Bell-Syer S, Jackson D, Llewlyn-Phillips H, Farrin A, et al.Randomised controlled trial of exercise for low back pain: clinical outcomes,costs, and preferences. British Medical Journal 1999;319(7205):279e83.

Nusbaum L, Natour J, Ferraz MB, Goldenberg J. Translation, adaptation and valida-tion of the RolandeMorris questionnaire e Brazil RolandeMorris. BrazilianJournal of Medical and Biological Research 2001;34(2):203e10.

O’Sullivan PB. Lumbar segmental ‘instability’: clinical presentation and specificstabilizing exercise management. Manual Therapy 2000;5(1):2e12.

O’Sullivan PB, Beales DJ, Beetham JA, Cripps J, Graf F, Lin IB, et al. Altered motorcontrol strategies in subjects with sacroiliac joint pain during the activestraight-leg-raise test. Spine 2002a;27(1):E1e8.

O’Sullivan PB, Grahamslaw KM, Kendell M, Lapenskie SC, Moller NE, Richards KV.The effect of different standing and sitting postures on trunk muscle activity ina pain-free population. Spine 2002b;27(11):1238e44.

O’Sullivan PB, Dankaerts W, Burnett AF, Farrell GT, Jefford E, Naylor CS, et al. Effect ofdifferent upright sitting postures on spinal-pelvic curvature and trunk muscleactivation in a pain-free population. Spine 2006;31(19):E707e12.

Panjabi MM. Clinical spinal instability and low back pain. Journal of Electromyog-raphy and Kinesiology 2003;13(4):371e9.

Pinto RZ, Ferreira PH, Franco MR, Ferreira ML, Ferreira MC, Teixeira-Salmela LF,et al. Effect of 2 lumbar spine postures on transversus abdominis musclethickness during a voluntary contraction in people with and without low backpain. Journal of Manipulative and Physiological Therapeutics 2011;34(3):164e72.

Richardson CA, Jull GA. Muscle controlepain control. What exercises would youprescribe? Manual Therapy 1995;1(1):2e10.

Ross RLP. Clinical assessment of pain. In: van Dieen JH, editor. Assessment inoccupational therapy and physical therapy. Philadelphia: WB Saunders; 1997. p.123e33.

Sapsford RR, Richardson CA, Stanton WR. Sitting posture affects pelvic floor muscleactivity in parous women: an observational study. Australian Journal of Phys-iotherapy 2006;52(3):219e22.

Siqueira FB, Teixeira-Salmela LF, Magalhães LC. Análise das propriedades psico-métricas da versão brasileira da escala Tampa de cinesiofobia. Acta OrtopédicaBrasileira 2007;15(1):19e24.

Springer BA, Mielcarek BJ, Nesfield TK, Teyhen DS. Relationships among lateralabdominal muscles, gender, body mass index, and hand dominance. Journal ofOrthopaedic and Sports Physical Therapy 2006;36(5):289e97.

Suni J, Rinne M, Natri A, Statistisian MP, Parkkari J, Alaranta H. Control of the lumbarneutral zone decreases low back pain and improves self-evaluated work ability:a 12-month randomized controlled study. Spine 2006;31(18):E611e20.

Teyhen DS, Bluemle LN, Dolbeer JA, Baker SE, Molloy JM, Whittaker J, et al. Changesin lateral abdominal muscle thickness during the abdominal drawing-inmaneuver in those with lumbopelvic pain. Journal of Orthopaedic and SportsPhysical Therapy 2009a;39(11):791e8.

Teyhen DS, Williamson JN, Carlson NH, Suttles ST, O’Laughlin SJ, Whittaker JL, et al.Ultrasound characteristics of the deep abdominal muscles during the activestraight leg raise test. Archives of Physical Medicine and Rehabilitation 2009b;90(5):761e7.

Tsao H, Hodges PW. Persistence of improvements in postural strategies followingmotor control training in people with recurrent low back pain. Journal ofElectromyography and Kinesiology 2008;18(4):559e67.

van Dieen JH, Cholewicki J, Radebold A. Trunk muscle recruitment patterns inpatients with low back pain enhance the stability of the lumbar spine. Spine2003;28(8):834e41.

van Tulder M, Koes B, Bombardier C. Low back pain. Best Practice and ResearchClinical Rheumatology 2002;16(5):761e75.

Waddell G. The back pain revolution. 2 ed. London: Churchill Livingstone; 2004.Westaway MD, Stratford PW, Binkley JM. The patient-specific functional scale:

validation of its use in persons with neck dysfunction. Journal of Orthopaedicand Sports Physical Therapy 1998;27(5):331e8.