www.elsevier.com/locate/clinbiomech

Clinical Biomechanics 19 (2004) 456–464

Altered patterns of pelvic bone motion determined in subjectswith posterior pelvic pain using skin markers

Barbara Hungerford a,*, Wendy Gilleard b, Diane Lee c

a School of Exercise and Sport Science, University of Sydney, Sydney, Australiab School of Exercise Science and Sport Management, Southern Cross University, Lismore, Australia

c Diane G. Lee Physiotherapist Corp., BC, Canada

Received 14 February 2003; accepted 19 February 2004

Abstract

Objective. To determine whether the pattern of pelvic bone motion, determined by skin markers, differs between control subjects

and subjects with posterior pelvic pain.

Design. Cross-sectional study of three-dimensional angular and translational motion of the innominates relative to the sacrum in

two subject groups.

Background. Comparative in vivo analysis of the 3D patterning of pelvic motion in subjects with posterior pelvic pain and

controls is limited.

Methods. Fourteen males with posterior pelvic pain and healthy age and height matched controls were studied. A 6-camera

motion analysis system was used to determine 3D angular and translational motion of pelvic skin markers during standing hip

flexion.

Results. Posterior rotation of the innominate occurred with hip flexion in control subjects and pelvic pain subjects as previously

reported in the literature. On the supporting leg, the innominate rotated posteriorly in controls and anteriorly in symptomatic

subjects.

Conclusion. Posterior rotation of the innominate, as measured using skin markers during weight bearing in controls may reflect

activation of optimal lumbo-pelvic stabilisation strategies for load transfer. Anterior rotation occurred in symptomatic subjects,

suggesting failure to stabilise intra-pelvic motion for load transfer.

Relevance

This study found that posterior rotation of the innominate occurred during weight bearing in controls. This movement pattern is

thought to optimise stability of the pelvic girdle during increased loading. Conversely, anterior rotation occurred in symptomatic

subjects during weight bearing. This is a non-optimal pattern and may indicate abnormal articular or neuromyofascial function

during increased vertical loading through the pelvis.

� 2004 Elsevier Ltd. All rights reserved.

Keywords: Sacroiliac joint; Pelvic motion; Pelvic stabilisation; Low back pain; Pelvic pain

1. Introduction

A primary function of the lumbar spine and pelvis is

to transfer the loads generated by body weight and

gravity during standing, walking and sitting (Snijders

et al., 1993). How well this load is managed dictates the

efficacy of function. A small amount of motion occurs at

* Corresponding author. Present address: Sydney Spine and Pelvis

Centre, 101 Lyons Road, Drummoyne, NSW 2047, Australia.

E-mail address: barbhungerford@aol.com (B. Hungerford).

0268-0033/$ - see front matter � 2004 Elsevier Ltd. All rights reserved.

doi:10.1016/j.clinbiomech.2004.02.004

the sacroiliac joints (SIJ) and the pubic symphysis dur-

ing movements of the trunk and lower limbs (Jacob and

Kissling, 1995; Walheim and Selvik, 1984). Conse-

quently, during weight bearing activities, control (sta-

bilisation) of intra-pelvic motion is required for

transference of loads between the spine and the lower

limbs (Snijders et al., 1993; Vleeming et al., 1990).According to Panjabi (1992) stability is achieved when

the passive, active and control systems work together.

Snijders et al. (1993) suggests that the passive, active and

control systems produce approximation of the joint

surfaces, essential if stability is to be insured. The

B. Hungerford et al. / Clinical Biomechanics 19 (2004) 456–464 457

amount of approximation required is variable and dif-ficult to quantify as it is dependent on an individual’s

structure (form closure) and the forces they need to

control (force closure). The ability to effectively transfer

load through the pelvis is dynamic and therefore de-

pends on: (1) optimal function of the bones, joints and

ligaments (Vleeming et al., 1989, 1990); (2) optimal

function of the muscles and fascia (Hungerford et al.,

2003; Richardson et al., 2002; Snijders et al., 1998;Vleeming et al., 1995a,b); (3) appropriate neural func-

tion (Hodges and Richardson, 1997; Hungerford et al.,

2003).

1.1. Stabilisation and motion control of the pelvic girdle

For every joint, there is a position called the self-

braced (close-packed) position in which there is maxi-mum congruence of the articular surfaces and maximum

tension on major ligaments. In this position, the joint is

under significant compression and the ability to resist

shear forces is enhanced by tensioning of the passive

structures and increased friction between the articular

surfaces (Snijders et al., 1993; Vleeming et al., 1990).

The self-braced position of the SIJ is nutation of the

sacrum or posterior rotation of the innominate (Vlee-ming et al., 1989). Studies have shown (Sturesson et al.,

2000) that nutation of the sacrum relative to posterior

rotation of the innominate occurs bilaterally whenever

the lumbo-pelvic spine is loaded vertically (sitting,

standing). Counternutation of the sacrum, or anterior

rotation of the innominate, is thought to be a relatively

less stable position for the SIJ (Vleeming et al., 1995b).

The long dorsal ligament becomes taut during thismotion, however tension in other ligaments such as

sacrotuberous and interosseous ligaments decreases

(Vleeming et al., 1996).

At present there is only limited research comparing

the in vivo pattern of pelvic motion during weight

bearing and non-weight bearing activities. This is due to

the invasiveness of the most reliable and valid methods

of analysis, consequential ethical considerations forusing invasive procedures to evaluate large number of

subjects and difficulty in evaluating the small amplitude

of in vivo SIJ motion. When an individual stands on

one leg and flexes the contralateral hip, the non-weight

bearing innominate posteriorly rotates relative to the

sacrum (range of motion 0.1–5.0�) (Jacob and Kissling,1995; Sturesson et al., 2000). Side flexion and axial

rotation of the innominate occur concurrently aboutsagittal and vertical axes respectively. Translation

(motion along the sagittal, vertical and/or coronal axes)

also occurs (Jacob and Kissling, 1995; Sturesson et al.,

2000), although the specific direction of translation that

occurs with angular motion is not fully understood. It

has been hypothesised (Lee, 1999) that anterior and

superior translation of the innominate, relative to the

sacrum, will occur with posterior rotation of theinnominate.

1.2. Range of motion versus patterning of pelvic motion

Buyruk et al. (1995) and Damen et al. (2002) estab-

lished that a Doppler imaging system was able to mea-

sure stiffness of the SIJ. This research showed that

stiffness of the SIJ is variable between subjects andtherefore the range of motion is potentially variable. It

also revealed that stiffness of right and left SIJs is sym-

metric in subjects without pelvic pain. The results sup-

port putting less emphasis on amplitude of motion and

more on the pattern or symmetry of SIJ, or pelvic

motion, as range of motion varies between subjects,

however within one subject joint stiffness remains sym-

metric between sides.

1.3. The impact of posterior pelvic pain

The SIJs and the posterior SIJ ligaments are a knownsource of posterior pelvic pain (Fortin et al., 1994;

Vleeming et al., 2002). Jacob and Kissling (1995) noted

that in the presence of SIJ symptoms, the amplitude of

SIJ motion about a coronal axis increased during hip

flexion in one subject. Mens et al. (1999) determined

increased amplitude of anterior rotation of the innom-

inate in posterior pelvic pain patients. Sturesson et al.

(2000) reported no difference in the amplitude or patternof either angular or translational motion of the in-

nominates when the left and right SIJs were compared

in subjects with posterior pelvic pain. No comparison

was made with an asymptomatic group. Hungerford

et al. (2003) showed that posterior pelvic pain alters the

pattern of lumbo-pelvic muscle recruitment, while

Buyruk et al. (1999) and Damen et al. (2002) showed

that stiffness of the SIJ is asymmetric in subjects withpelvic pain and that asymmetrical stiffness of the SIJs is

prognostic for pelvic impairment and pain. It is pres-

ently unknown if posterior pelvic pain alters the pat-

terning of bone motion within the pelvis for single leg

stance.

1.4. Posterior pelvic pain and the active straight leg raise

test

The supine active straight leg raise test (ASLR)

(Mens et al., 2001) has been validated as a clinical test

for measuring effective load transfer between the trunk

and lower limbs. When the lumbo-pelvic region is

functioning optimally, the leg should rise effortlessly

from the table (effort graded from 0 to 5) (Mens et al.,

1999). A correlation has been shown between positiveASLR findings and posterior pelvic pain (Mens et al.,

1999; O’Sullivan et al., 2002). Similarly, Damen et al.

458 B. Hungerford et al. / Clinical Biomechanics 19 (2004) 456–464

(2001) and Buyruk et al. (1999) showed that the ASLRis positive in the presence of asymmetric stiffness of the

SIJ. This suggests that altered pelvic stabilisation strat-

egies may affect pelvic mobility (Buyruk et al., 1999;

Damen et al., 2001).

1.5. Limitations of methodology in this study

A non-invasive motion analysis system using skinmounted markers was chosen to acquire the kinematic

data of pelvic bone motion during a standing hip flexion

movement for ethical reasons. Errors in determining the

range of motion are likely to occur with an opto-

electronic system due to skin marker motion relative to

underlying bony landmarks (Maslen and Ackland,

1994). A high resolution motion analysis system has

however been reported to provide reliable and consistentin vivo data of lumbar segmental motion patterns

(Gracovetsky et al., 1995). In this study, the authors

recognise that the movements noted reflect motion of

the innominates, sacrum, and femurs in conjunction

with overlying skin, and therefore the main emphasis of

this study was to investigate the patterns of bone motion

rather than the range of motion.

The aim of this study was to determine the three-dimensional pattern of innominate bone motion occur-

ring in subjects determined to have posterior pelvic pain

and impaired pelvic stabilisation strategies during

weight bearing and non-weight bearing components of a

standing hip flexion movement. These results were

compared to age and height matched controls with

clinically assessed normal pelvic stabilisation and pelvic

motion patterns. It was hypothesised that the pattern ofinnominate bone motion would alter in subjects with

posterior pelvic pain during both components of the

movement trial.

2. Methods

2.1. Subjects

2.1.1. Impaired pelvic stabilisation and posterior pelvic

pain group

Fourteen male subjects with SIJ pain and a mean

(range) age, height and weight of 32.7 (24–47) years,

176.8 (168–184) cm, and 77.0 (71–90) kg respectively,

volunteered for the study. The criteria for inclusion in

this study were:

1. Each subject in the posterior pelvic pain group

reported unilateral pain over the posterior pelvic/SI

region (Fortin et al., 1994) for greater than two

months, and no pain above the lumbo-sacral junc-

tion. The pain was consistently and predictably

aggravated by activities that vertically loaded the pel-vis (walking, standing or sitting).

2. Positive results on the side of posterior pelvic pain in

clinical tests for impaired lumbo-pelvic stabilisation.

These tests included:

(a) Active straight leg raise test (Mens et al., 1999,

2001): A positive test was indicated when the pelvis

failed to remain in neutral alignment, and the sub-

ject reported difficulty or inability to elevate astraight leg in supine. The perceived difference of

effort, or pain aggravation was scaled from 0

(not difficult to raise the leg) to 5 (unable to per-

form ASLR).

(b) Standing hip flexion test (Mitchell, 1995): Dur-

ing a left standing hip flexion test, the subject

stands on their right leg and flexes the left hip to-

wards 90�. The left innominate should posteriorlyrotate relative to the sacrum (Jacob and Kissling,

1995). A positive test was indicated when superior

motion of the posterior superior iliac spine (PSIS)

was palpated relative to the sacrum.

(c) Neutral zone analysis test (joint play) (Lee,

1999)––this test was used clinically to apply the re-

search of Buyruk et al. (1999) and Damen et al.

(2002) and to evaluate motion in the neutral zoneof the SIJ. Panjabi (1992) noted that joints have

non-linear load–displacement curves and that the

size of the neutral zone may increase with injury,

articular degeneration and/or weakness of the sta-

bilising musculature and that this is a more sensi-

tive indicator than angular range of motion for

detecting instability. All symptomatic subjects

demonstrated asymmetric stiffness of the SIJ whenthe innominate was glided relative to the sacrum

(analysis of the neutral zone).

As the reliability and predictive ability of the standing

hip flexion and neutral zone analysis tests remains

uncertain (Carmichael, 1987; Vincent-Smith and Gib-

bons, 1999) all clinical tests were required to be positive

on the side of pain, in conjunction with the ASLR test,for inclusion in the posterior pelvic pain subject group.

Subjects were excluded from the study if they could not

flex each hip to 90� without pain, if they had undergonespinal surgery, or displayed overt neurological signs

such as sensory paraesthesia or motor paresis.

2.1.2. Control group

The SIJ pain group were age and height matched to acontrol group of 14 males with a mean (range) age,

height, and weight of 33.5 (22–50) years, 176.0 (168–183)

cm, and 72.5 (61–85) kg respectively. The control sub-

jects had no history of low back pain in the last 12

months, no history of congenital lumbar or pelvic

anomalies, and tested negative to the ASLR and

standing hip flexion and neutral zone analysis tests.

Fig. 1. The three axes for angular and translational motion of the

innominate relative to the sacral segment. Note the axis of innominate

segment motion is centred at the PSIS.

B. Hungerford et al. / Clinical Biomechanics 19 (2004) 456–464 459

Subjects were excluded if no palpable motion betweenthe PSIS and S2 spinous process was observed during

unilateral hip flexion in standing, or if they experienced

any pain during the clinical assessment.

All subjects were assessed by the same experienced

physiotherapist to maintain continuity. Informed con-

sent was given by each subject prior to participation in

the study, and all rights of the subjects were protected.

The study was approved by the institutional HumanEthics Committee.

2.2. Procedure

Fifteen lightweight highly reflective 15 mm diameter

balls were used to define the bony landmarks of each

innominate, both femoral segments, and the sacrum.

The pelvic bony landmarks were chosen for theircloseness to the skin surface with minimal overlying

fascia, and because they reflected palpation points

commonly used by therapists. Each innominate was

therefore defined by three markers placed on the ante-

rior superior iliac spine (ASIS), the PSIS, and the lateral

iliac tubercle. A three armed triangular wand with a

single marker attached to each arm was applied to the

sacral spinous process of S2. Each arm was 10 mm long,and solidly welded to the triangular base. The apical

marker of the wand approximated the S1 spinous pro-

cess, while a left and right marker formed the base of the

triangle at the horizontal level of S2. The left and right

femoral segments were defined by markers on the

greater trochanter, the lateral femoral condyle, and a

mid thigh marker placed 20 cm inferior and 5 cm

anterior to the greater trochanter. All markers wereapplied to the skin while the subjects were standing.

A six camera Expert Vision Motion Analysis (Eva)e

Hi Res.6.0 System (Motion Analysis Corporation, Cal-

ifornia, USA) was used to video (60 Hz) the subject

motion. Measurements of a known angle showed the six

camera Evae system was accurate to 0.25�. Forceplatform data (960 Hz) were used to identify initiation

of single leg support during each standing hip flexiontrial. Following practice trials, data from one quiet

standing trial were collected. The subject then per-

formed six left and six right standing hip flexion trials.

For example, during a left standing hip flexion trial the

subject was asked to stand on his right leg, and flex his

left hip and knee toward 90� hip flexion, then lower thefoot back down.

2.3. Data analysis

The Evae motion analysis system was used to track

the 3D trajectories of each marker over time. These

trajectories were then imported into Kintrake (Motion

Analysis Corporation) which provided the 3D angular

rotation, and translation, of each innominate relative to

the sacral segment throughout each trial, in respect to

neutral position from the quiet standing trial. Calcula-tion of the angular kinematics required definition of

individual segment coordinate systems, and a joint

coordinate system. The segment coordinate systems

were assumed to be embedded within each adjacent

segment. The axes of the innominate segment originated

at each respective PSIS, and the axes of the sacral seg-

ment intercepted at the S2 spinous process (Fig. 1).

Pelvic motion was determined by aligning the coronalaxis of each innominate and sacral segment to intersect

both the PSIS and S2 spinous process. Kintrake sub-

sequently determined relative angular and translational

motion of each bone segment about their aligned seg-

ment coordinate systems. Hip joint motion was deter-

mined by computing the motion of the femur, as defined

by linking the three femoral markers, in relation to

ipsilateral innominate motion. The hip joint centre wasdetermined using the equation provided by Tylkowski

et al. (1982).

All angular and translational motion was determined

at maximum coronal axis motion of the innominate,

relative to the sacrum, during hip flexion. Two tailed

paired Student t-tests assuming unequal variance

(Domholdt, 1993) were performed for all variables be-

tween the left and right side in the control group, andbetween the symptomatic side and the asymptomatic

side in the SIJ pain group. In order to determine if there

was a significant alteration to the mean angular and

translational intra-pelvic motion measured between the

control subjects and the SIJ pain group, independent

groups two tailed Student t-tests assuming unequal

variance (Domholdt, 1993) were performed for all

variables. Further graphical comparisons of coronalaxis motion on the side of single leg support were

-6

-4

-2

0

2

4

6

8

10

12

1 2 3 4 5 6 7 8 9 10 11 12 13 14

subjects

coronalaxisrotation(°)

posterior

anterior

pelvic pain asymptomaticpelvic pain symptomaticcontrols mean

Fig. 2. A comparison of coronal axis angular motion at single leg support in each subject with posterior pelvic pain. Mean coronal axis motion for

control subjects is depicted as a solid line. Error bars denote 2 · standard error for each subject. �Significant at P < 0:05.

460 B. Hungerford et al. / Clinical Biomechanics 19 (2004) 456–464

performed for each subject in the SIJ pain group. Two

times the standard error of mean motion for eachsubject was determined and plotted. As the mean± 2 SE

is approximately equal to the 95% confidence interval

(Sim and Reid, 1999), determination of no overlay of

each subject’s data was defined as a significant difference

(Fig. 2).

3. Results

The pattern of angular and translational motion of

each innominate segment, relative to the sacral segment,

Table 1

Comparison of angular and translational motion of the left and right innom

posterior pelvic pain

Motion

segment

Axis of motion A: Control subjects

Left hip flexion Right hip flexion

Mean SD Mean SD

Femur Coronal 70.00 5.25 73.00 5.50

Innominate

angular (�)Coronal )8.50 3.50 )10.00 3.50

Sagittal )6.00 3.50 )7.75 5.00

Vertical 4.50 4.50 3.50 3.00

Innominate

translation

(mm)

Antero-posterior 3.50 2.50 4.00 3.00

Medio-lateral )5.50 3.00 )5.50 2.50

Vertical )6.50 3.00 )7.50 3.50

Negative coronal value¼ posterior; negative sagittal¼ toward flexed hip; neganegative medio-lateral translation¼ toward flexed hip; negative vertical tran* Significant at P < 0:05.

was measured during a standing hip flexion movement

using a non-invasive method. Possible errors in deter-mining the magnitude of pelvic bone motion may have

occurred due to the movement of skin markers over

bony landmarks, therefore the emphasis of this study

was on changes to the patterns of pelvic bone motion as

reflected by the measured skin marker movement.

3.1. Hip flexion side

The angular and translational motions of the left and

right innominates, during left and right standing hip

flexion movements are summarised in Tables 1–3. In the

inates on the side of hip flexion in control subjects and subjects with

B: Posterior pelvic pain subjects

P -value Asymptomatic

hip flexion

Symptomatic hip

flexion

P -value

Mean SD Mean SD

0.08 73.25 7.00 74.50 4.50 0.28

0.05 )7.50 3.00 )10.00 3.25 0.04�

0.18 )5.00 5.50 )6.25 4.00 0.44

0.39 3.75 2.50 4.75 1.75 0.27

0.49 2.00 4.00 2.00 3.75 0.70

0.74 )5.50 2.00 )6.50 3.00 0.36

0.46 )4.00 3.50 )5.00 2.00 0.76

tive vertical¼ toward flexed hip; negative antero-posterior¼ posterior;slation¼ superior.

Table 2

Comparison of angular and translational motion of the innominates on the side of single leg support in control subjects and subjects with posterior

pelvic pain

Motion

segmentAxis of motion A: Control subjects B: Posterior pelvic pain subjects

Left single leg

support

Right single leg

support

P -value Asymptomatic

single leg support

Symptomatic

single leg support

P -value

Axis Mean SD Mean SD Mean SD Mean SD

Femur Coronal 1.50 3.00 1.50 2.50 0.79 1.75 3.50 1.75 3.25 0.86

Innominate

angular (�)Coronal 0.00 2.00 )0.50 3.00 0.88 0.50 1.50 2.00 2.00 0.02�

Sagittal )6.00 2.50 )5.75 2.50 0.46 )4.75 3.25 )3.75 3.00 0.42

Vertical 3.50 2.75 4.50 4.50 0.17 5.00 2.50 3.75 2.75 0.09

Innominate

translation

(mm)

Antero-posterior )4.50 6.00 )2.50 4.00 0.13 )4.00 2.50 )6.00 2.75 0.02�

Medio-lateral )6.50 3.25 )6.50 3.00 0.96 )7.00 3.25 )5.00 2.50 0.03�

Vertical )4.50 2.50 )3.50 2.50 0.19 3.50 2.25 2.00 3.00 0.04�

Negative coronal value¼posterior; negative sagittal¼ toward flexed hip; negative vertical¼ toward flexed hip; negative A-P translation¼ posterior;negative medio-lateral¼ toward flexed hip; negative vertical¼ superior.

Table 3

Comparison of angular and translational motion of the innominates between control subjects and subjects with posterior pelvic pain

Motion

segmentAxis of motion A: Hip flexion B: Single leg support

Controls right Pelvic pain group

symptomatic

P -value Controls right Pelvic pain group

symptomatic

P -value

Axis Mean SD Mean SD Mean SD Mean SD

Femur Coronal 73.00 5.50 74.50 4.50 0.45 1.50 2.50 1.75 3.25 0.67

Innominate

angular (�)Coronal )10.00 3.50 )10.00 3.25 0.16 )0.50 3.00 2.00 2.00 <0.01�

Sagittal )7.75 5.00 )6.25 4.00 0.94 )5.75 2.50 )3.75 3.00 0.12

Vertical 3.50 3.00 4.75 1.75 0.46 4.50 4.50 3.75 2.75 0.90

Innominate

translation

(mm)

Antero-posterior 4.00 3.00 2.00 3.75 0.05 )2.50 4.00 )6.00 2.75 0.02�

Medio-lateral )5.50 2.50 )6.50 3.00 0.94 )6.50 3.00 )5.00 2.50 0.13

Vertical )7.50 3.50 )5.00 2.00 0.04� )3.50 2.50 2.00 3.00 <0.001�

Negative coronal value¼posterior; negative sagittal¼ toward flexed hip; negative vertical¼ toward flexed hip; negative A-P translation¼ posterior;negative medio-lateral¼ toward flexed hip; negative vertical¼ superior.* Significant at P < 0:05.

B. Hungerford et al. / Clinical Biomechanics 19 (2004) 456–464 461

control subjects, standing hip flexion (Table 1A) pro-

duced posterior rotation of the non-weight bearinginnominate relative to the sacral segment. On the side of

hip flexion, femoral flexion showed a mean (range) 73.0�(10.0) at maximum coronal axis angular motion of the

innominate. The innominate concurrently side flexed

toward the side of hip flexion, and rotated about the

vertical axis away from the side of hip flexion. Trans-

lation of the innominate relative to the sacral segment

was associated with this angular motion. The innomi-nate translated anteriorly, superiorly, and laterally (to-

ward the side of hip flexion) as the innominate

posteriorly rotated (Table 1A). No significant difference

was found in the controls between the pattern of left and

right innominate motion on the side of hip flexion.

On the side of hip flexion for the posterior pelvic pain

group, maximum posterior rotation of the innominate

occurred at a mean of 73.25� (7.0) femoral flexion on theasymptomatic side, and at a mean of 74.5� (4.5) femoralflexion on the symptomatic side (Table 1B). A pattern of

side flexion of the innominate, and rotation about the

vertical axis away from the side of hip flexion, occurredon both the asymptomatic and symptomatic sides dur-

ing hip flexion.

Anterior, superior, and lateral translation of the

innominate toward the side of hip flexion, were associ-

ated with posterior rotation of the innominate on the

side of hip flexion. In the posterior pelvic pain group

there was no significant difference between the asymp-

tomatic and symptomatic sides in translational motion,or angular motion of the innominate about the sagittal

or vertical axes during hip flexion (Table 1B).

3.2. Single leg support

The contralateral limb maintained single leg support

during the standing hip flexion movement. Posterior

rotation of the weight bearing innominate occurredabout the coronal axis in control subjects (Table 2A). A

concurrent pattern of side flexion of the innominate

462 B. Hungerford et al. / Clinical Biomechanics 19 (2004) 456–464

toward, and rotation about the vertical axis away fromthe side of hip flexion also occurred. The weight bearing

innominate translated posteriorly, superiorly and

medially (toward the side of hip flexion) (Table 2A). No

significant difference was found in the pattern of either

angular or translational motion of the weight bearing

innominate during single leg support in control subjects.

The angular and translational motion of the weight

bearing innominate on the side of single leg support, insubjects with posterior pelvic pain, is depicted in Table

2B. The weight bearing innominate anteriorly rotated

significantly more (P ¼ 0:02) about the coronal axis onthe symptomatic in comparison to the asymptomatic

side. Concurrently, the innominate side flexed toward,

and rotated away from the side of hip flexion on both

the symptomatic and asymptomatic sides. The innomi-

nate translated inferiorly, posteriorly, and medially(toward the side of hip flexion) relative to the sacral

segment (Table 2B). Posterior translation of the

innominate relative to the sacral segment was signifi-

cantly greater (P ¼ 0:02) on the symptomatic side thanthe asymptomatic side, while inferior translation of the

innominate was significantly less (P ¼ 0:04) on the

symptomatic side in comparison to the asymptomatic

side. Medial translation, or compression, of theinnominate was also significantly less (P ¼ 0:03) on thesymptomatic side in comparison to the asymptomatic

side (Table 2B).

Further comparison of coronal axis motion on the

symptomatic side and asymptomatic side of single leg

support was performed for each subject with posterior

pelvic pain (Fig. 2). The mean range of posterior rota-

tion on the side of single leg support in control subjectswas depicted as a solid line for comparison to the

symptomatic group. A significant change toward ante-

rior rotation of the innominate occurred in 12 of the 14

posterior pelvic pain subjects on the symptomatic side;

that is in subjects 1, 2, 4 to 7, 9 to 14 (Fig. 2).

3.3. Comparison of control subjects and subjects with

posterior pelvic pain

During the hip flexion component of the standing hip

flexion movement there was no significant difference in

patterning of angular or translational motion between

groups on the side of hip flexion (Table 3A).

On the side of single leg support, a significant differ-

ence in the pattern of angular and translational motion

of the innominate between controls and symptomaticsubjects was determined (Table 3B). In the control

group, posterior rotation of the weight bearing innom-

inate occurred on the side of single leg support; however,

in the posterior pelvic pain group, anterior rotation

occurred on the symptomatic side (P < 0:01). A pattern

of posterior, superior and medial translation occurred

concurrently with posterior rotation of the innominate

in control subjects; however, inferior, posterior, andmedial translation occurred with anterior rotation of the

innominate in the symptomatic subjects.

4. Discussion

Significant differences were found in the pattern of

pelvic bone motion that occurred during standing hipflexion when intra-subject (between weight bearing and

non-weight bearing sides) and inter-subject (between the

control group and posterior pelvic pain group) com-

parisons were made. The range of motion reported be-

tween the innominate and sacral segments, as

determined using motion analysis, was generally larger

than the range of motion determined during analysis of

SIJ motion using stereophotogrammetry (Sturessonet al., 2000). This may be due to errors created by skin

deformation over bone landmarks, or the movement of

muscles close to the placement of the skin markers, such

as gluteus maximus or obliquus abdominis internus, as

they activated to create the test movement (Hungerford

et al., 2003).

4.1. Intra-subject comparisons––control group

In the control subjects hip motion toward 90� offemoral flexion produced posterior rotation of the non-

weight bearing innominate (side of hip flexion), consis-

tent with previous research (Jacob and Kissling, 1995;

Sturesson et al., 2000). A concurrent pattern of side

flexion toward the side of hip flexion, and rotation about

the vertical axis away from the side of hip flexion wasfound. In addition, a concurrent translational motion

occurred (anterior, superior and lateral) during hip

flexion. This pattern of translation is consistent with the

model of arthrokinematic motion of the SIJ proposed

by Lee (1999).

During this same movement, the weight bearing

innominate (side of single leg support) posteriorly ro-

tated; a finding consistent with previous research(Sturesson et al., 2000). In addition, this study found a

concurrent side flexion of the innominate toward, and

axial rotation away from the side of hip flexion. Pos-

terior rotation of the innominate (or sacral nutation) is

thought to occur as a consequence of the self-bracing

mechanism of the pelvis; essential for optimal load

transfer during single leg support (Snijders et al., 1993;

Vleeming et al., 1989, 1995b). A concurrent translationalmotion occurred (posterior, superior and medial) during

single leg support. Although both the non-weight bear-

ing and weight bearing innominates posteriorly rotated

during standing hip flexion, the pattern of the concur-

rent translation between the innominate and sacral

segments on the non-weight bearing and weight bearing

sides differed. This variation of translational motion

B. Hungerford et al. / Clinical Biomechanics 19 (2004) 456–464 463

may reflect different patterns of muscle activation anddifferent compressive forces acting on the innominate

during single leg support. Using the Doppler imaging

system, Richardson et al. (2002) noted that a co-

contraction of multifidus and transversus abdominis

increased the stiffness of the SIJ. An optimal lumbo-

pelvic stabilisation strategy requires recruitment of these

muscles and could explain the compression (medial

translation of the innominate) noted in the controlsubjects during single leg support.

4.2. Intra-subject comparisons––posterior pelvic pain

group

During hip flexion in the subjects with posterior pelvic

pain, hip motion toward 90� of femoral flexion also

produced posterior rotation of the non-weight bearinginnominate. This pattern of motion was found on both

the symptomatic and asymptomatic side. The pattern of

side flexion, rotation, and translation of the non-weight

bearing innominate did not differ between the asymp-

tomatic and symptomatic sides.

On the side of single leg support, anterior rotation of

the weight bearing innominate occurred in subjects with

posterior pelvic pain. This pattern of motion has beennoted previously by Mens et al. (1999) during single leg

loading in patients with posterior pelvic pain. In addi-

tion, this study found that the concurrent translational

motion of the weight bearing innominate differed in

pelvic pain patients. As the innominate rotated anteri-

orly it translated inferiorly and posteriorly; a pattern

hypothesised to be consistent with less intra-pelvic

compression (Lee, 1999). It is interesting to note that inthis study, all of the subjects in the posterior pelvic pain

group reported increased symptoms with vertical load-

ing through the pelvis (standing, walking). This may

suggest they were unable to adequately compress the

weight bearing SIJ and maintain self-bracing of the

pelvis (Snijders et al., 1998) in order to control vertical

shear loads.

4.3. The two groups––inter-subject comparisons

A comparison of both the angular and translational

motion of the non-weight bearing innominate on the

side of hip flexion between control subjects and matched

subjects with posterior pelvic pain showed no significant

difference. This has important clinical relevance. Clinical

assessment of posterior rotation of the innominate onthe side of hip flexion (the stork test) as a method of

distinguishing normal joint motion from SIJ dysfunc-

tion (Mitchell, 1995) has been found to be unreliable

and unspecific (Vincent-Smith and Gibbons, 1999). This

study further validates such conclusions.

A significant difference was noted in the pattern of

angular motion of the weight bearing innominate during

single leg support when the control group (posteriorrotation) was compared to the posterior pelvic pain

group (anterior rotation). Vleeming et al. (1995b) sug-

gest that anterior rotation of the innominate (sacral

counternutation) disengages the self-bracing mechanism

of the pelvis and consequently diminishes the ability to

transfer loads between the spine and legs. The positive

clinical tests noted in the posterior pelvic pain group

may reflect their inability to posteriorly rotate theweight bearing innominate on the symptomatic side.

Alternately, subjects may have lacked the ability to

create sufficient compression due to altered motor con-

trol of the musculature known to stabilise the pelvis,

such as transversus abdominis, lumbo-sacral multifidus,

and gluteus maximus (Hungerford et al., 2003; Rich-

ardson et al., 2002).

5. Conclusions

The most significant alteration to the pattern of bone

motion between controls and subjects with posterior

pelvic pain occurred on the side of single leg support. In

the control subjects, the weight bearing innominate

posteriorly rotated and translated superiorly, posteriorlyand medial relative to the sacral segment. In the subjects

with posterior pelvic pain, the weight bearing innomi-

nate anteriorly rotated and translated inferiorly. The

results of this study suggest that posterior rotation of the

innominate is a normal component for optimal stabili-

sation of the pelvis. Anterior rotation of the innominate

is indicative of failure of the self-bracing mechanism and

load transfer through the pelvis, with a resultant de-crease in the ability to oppose vertical shear loads during

weight bearing.

Acknowledgements

The technical assistance of Ray Patton and Dr.

Richard Smith from the School of Exercise and Sports

Science, University of Sydney is acknowledged.

References

Buyruk, H.M., Snijders, C.J., Vleeming, A., Lam�eris, J.S., Holland,

W.P.J., Stam, H.J., 1995. The measurements of sacroiliac joint

stiffness with colour Doppler imaging: a study on healthy subjects.

European Journal of Radiology 21, 117–121.

Buyruk, H.M., Snijders, C.J., Vleeming, A., Lameris, J.S., Holland,

W.P.J., Stam, H.J., 1999. Measurements of sacroiliac joint stiffness

in peripartum pelvic patients with Doppler imaging of vibrations.

European Journal of Radiology 83, 159–163.

Carmichael, J., 1987. Inter and intra-examiner reliability of palpation

of the sacroiliac joint dysfunction. Journal of Manipulative and

Physiological Therapeutics 10, 164–171.

464 B. Hungerford et al. / Clinical Biomechanics 19 (2004) 456–464

Damen, L., Buyruk, H.M., Guler-Ulysal, F., Lotgering, F.K., Snijders,

C.J., Stam, H.J., 2001. Pelvic pain during pregnancy is associated

with asymmetric laxity of the sacroiliac joints. Acta Obstetricia et

Gynecologica Scandinavica 80, 1019–1024.

Damen, L., Buyruk, H.M., Guler-Uysal, F., Lotgering, F.K., Snijders,

C.J., Stam, H.J., 2002. Prognostic value of asymmetric laxity of the

sacroiliac joints in pregnancy-related pelvic pain. Spine 27, 2820.

Domholdt, E., 1993. Physical Therapy Research: Principles and

Applications. W.B Saunders Co, Philadelphia.

Fortin, J.D., Dwyer, A., West, S., Pier, J., 1994. Sacroiliac joint

referral patterns upon application of a new injection/arthrogra-

phy technique. I: Asymptomatic volunteers. Spine 19, 1475–

1482.

Gracovetsky, S., Newman, N., Pawlowsky, M., Lanzo, V., Davey, B.,

Robinson, L., 1995. A database for estimating normal spinal

motion derived from noninvasive measurements. Spine 20, 1036–

1046.

Hodges, P.W., Richardson, C.A.., 1997. Contraction of the abdominal

muscles associated with movement of the lower limb. Physical

Therapy 77, 132–144.

Hungerford, B., Gilleard, W., Hodges, P.W., 2003. Evidence of altered

lumbo-pelvic muscle recruitment in the presence of sacroiliac joint

pain. Spine 28, 1593–1600.

Jacob, H., Kissling, R., 1995. The mobility of the sacroiliac joints in

healthy volunteers between 20 and 50 years of age. Clinical

Biomechanics 10, 352–361.

Lee, D., 1999. The Pelvic Girdle: An Approach to Examination and

Treatment of the Lumbo-Pelvic-Hip Region. Churchill Living-

stone, Edinburgh.

Maslen, B., Ackland, T., 1994. Radiographic study of skin displace-

ment errors in the foot and ankle during standing. Clinical

Biomechanics 9, 291–296.

Mens, J.M.A., Vleeming, A., Snijders, C.J., Stam, H.J., Ginai, A.Z.,

1999. The active straight leg raising test and mobility of the pelvic

joints. European Spine Journal 8, 468–473.

Mens, J.M., Vleeming, A., Snijders, C.J., Koes, B., Stam, H.J., 2001.

Reliability and validity of the active straight leg raise test in

posterior pelvic pain since pregnancy. Spine 26, 1167–1171.

Mitchell, F.J., 1995. The Muscle Energy Manual, vol. 1. MET Press,

East Lansing.

O’Sullivan, P.B., Beales, D.J., Beetham, J.A., et al., 2002. Altered

motor control strategies in subjects with sacroiliac joint pain during

the active straight leg raise test. Spine 27, E1–E8.

Panjabi, M.M., 1992. The stabilising system of the spine. Part 2:

Neutral zone and stability hypothesis. Journal of Spinal Disorders

5, 390–397.

Richardson, C.A., Snijders, C.J., Hides, J.A., Damen, L., Martijn,

S.P., Storm, J., 2002. The relationship between transversus

abdominis muscles, sacroiliac joint mechanics, and low back pain.

Spine 27, 399–405.

Sim, J., Reid, N., 1999. Statistical inference by confidence intervals:

issues of interpretation and utilization. Physical Therapy 79, 186–

195.

Snijders, C.J., Ribbers, M.T., de Bakker, H.V., Stoeckart, R., Stam,

H.J., 1998. EMG recordings of abdominal and back muscles in

various standing postures: validation of a biomechanical model on

sacroiliac joint stability. Journal of Electromyography and Kine-

siology 8, 205–214.

Snijders, C.J., Vleeming, A., Stoeckart, R., 1993. Transfer of lumbo-

sacral load to iliac bones and legs. 1: Biomechanics of self-bracing

of the sacroiliac joints and its significance for treatment and

exercise. Clinical biomechanics 8, 285–294.

Sturesson, B., Uden, A., Vleeming, A., 2000. A radiological analysis of

movements of the sacroiliac joint during the standing hip flexion

test. Spine 25, 364–368.

Tylkowski, C.M., Simon, S.R., Mansour, J.M., 1982. Internal rotation

gait in spastic cerebral palsy in the hip. Paper presented at the

Proceedings of the 10th Open Scientific Meeting of the Hip Society,

St Louis. pp. 89–125.

Vincent-Smith, B., Gibbons, P., 1999. Inter-examiner and intra-

examiner reliability of the standing hip flexion test. Manual

Therapy 4, 87–93.

Vleeming, A., de Vries, H.J., Mens, J.M., van Wingerden, J.P., 2002.

Possible role of the long dorsal sacroiliac ligament in women with

peripartum pelvic pain. Acta Obstetrics Gynecology Scandinavia

81, 430.

Vleeming, A., Pool-Goudzwaard, A.L., Hammudoghlu, D., Stoeckart,

R., Snijders, C.J., Mens, J.M., 1996. The function of the long

dorsal sacroiliac ligament: its implication for understanding low

back pain. Spine 21, 556–562.

Vleeming, A., Pool-Goudzwaard, A.L., Stoeckart, R., van Wingerden,

J.P., Snijders, C.J., 1995a. The posterior layer of the thoracolum-

bar fascia: its function in load transfer from spine to legs. Spine 20,

753–758.

Vleeming, A., Snijders, C.J., Stoeckart, R., Mens, J.M., 1995b. A new

light on low back pain. Paper presented at the 2nd Interdisciplinary

World Congress on Low Back Pain, San Diego, pp. 149–168.

Vleeming, A., van Wingerden, J.P., Snijders, C.J., Stoeckart, R.,

Stijnen, T., 1989. Load application to the sacrotuberous ligament:

influences on sacroiliac joint mechanics. Clinical Biomechanics 4,

204–209.

Vleeming, A., Volkers, A.C.W., Snijders, C.J., Stoeckart, R., 1990.

Relation between form and function in the sacroiliac joint. 2:

Biomechanical aspects. Spine 15, 133–136.

Walheim, G.G., Selvik, G., 1984. Mobility of the pubic symphysis.

Clinical Orthopaedics and related Research 191, 129–135.