2.1    Relevant trunk functional anatomy

Figure 2.1 The long erector spinae muscles.
Schematic representation of thoracic portions of erector spinae. A. Fascicles of the right thoracic part of longissimus thoracis = longissimus thoracis pars thoracis. B. Fascicles of the right thoracic portion of iliocostalis lumborum = iliocostalis lumborum pars thoracis (adapted from Macintosh and Bogduk 1987).

The long posterior muscles of the trunk attach to the pelvis and thorax; that is they span over the lumbar spine (Macintosh and Bogduk 1987, 1991). Originating from the pelvis and some lumbar vertebrae, there is extensive insertion of fascicles to the ribs for both groups of long muscles (Figure 2.1). Sagittal and frontal standing motions of the trunk rely on these attachments for leverage and control of the thorax over the lumbar spine and pelvis. These long erector spinae muscles have been described as being suited to contributing to trunk moments and movements while the shorter lumbar muscles are better suited to control shear forces (Callaghan and McGill 1995).
From modelling of the back muscles acting during trunk extension from flexion, about 50% of the maximum extensor moment exerted on L5-S1 and 70-80% for the upper lumbar levels is produced by the long muscles acting from the pelvis to the thorax (thoracic fibres of longissimus thoracis and iliocostalis lumborum) (Bogduk et al 1992). The multifidus and the lumbar fascicles of longissimus and iliocostalis attaching only to the lumbar spine (and pelvis) control the anterior sagittal rotation of the lumbar vertebrae and contribute to the remaining 50% of the maximum extensor moment at L5-S1 (Bogduk et al 1992).
Anteriorly, the abdominal muscles arise from the front of the pelvis and attach to the thorax (Kendall and McCreary 1983) again bypassing any lumbar spinal attachment. Notably, the external oblique muscle originates from the anterolateral surface of the fifth rib through to the twelfth, hence its action pulls on eight ribs to affect trunk rotation and flexion (Kendall and McCreary 1983) (Figure 2.2). Again anatomical relationships define functional relationships between the pelvis, lumbar spine and thorax.

These significant anatomical and functional features of trunk musculature suggest that mobility and motor control of the trunk is related to the mobility and motor control of the thorax relative to the pelvis. The deep short lumbar muscles combine with the long erector spinae muscles to control trunk and hence lumbar spine motion. That one anterior trunk muscle arises from eight ribs and that one set of the posterior trunk muscles attaches to all 12 ribs suggests some role for the thorax in motor control, mobility and strength for trunk actions.
Trunk muscle behaviour during physiological movements (flexion, extension and lateral flexion) has been investigated extensively using EMG. With slow flexion, the erector spinae muscles decrease their tonic activity, initiating a fall forwards according to gravitational forces (Oddsson 1990). The movement is then controlled by the lengthening extensors (Floyd and Silver 1955; Thorstensson et al 1985; Crenna et al 1987). As forward flexion proceeds, a critical point near full flexion is reached where the lumbar extensors may become silent (Morris et al 1962; Kippers and Parker 1985) but not the thoracic extensors (McGill and Kippers 1994).

Figure 2.2 External oblique and rectus abdominus.
Schematic representation of the origin, insertion and line of action for the right external oblique and rectus abdominus muscles (adapted from Kendall and McCreary 1983).

Complete relaxation of erector spinae muscles near full flexion has been defined as the “flexion relaxation response” (Floyd and Silver 1950; Morris et al 1962; Kippers and Parker 1984; Ortengren and Andersson 1977; Bogduk et al 1992). Similarly, Basmajian and DeLuca (1985) also found the multifidi and rotators muscles to relax in full flexion.
Using fine wire electrodes, Andersson et al (1996) differentiated between the deep short lumbar erector spinae muscles (lumbar fascicles of longissimus and iliocostalis) and the long superficial muscles (thoracic fascicles of longissimus and iliocostalis) during trunk flexion. These authors reported relaxation in the long superficial muscles but not the deep muscles. Quadratus lumborum also remained active at full flexion. Hence at full flexion the spine is supported by some muscular control and not only by its ligaments as previously reported (Basmajian and DeLuca 1985; Bogduk et al 1992). McGill and Kippers (1994) found the erector spinae in the thoracic region to remain active while the lumbar muscles relaxed during full trunk flexion. Presumably, the short erector spinae muscles relaxed while the long muscles remained active.

This critical point is not always present (Floyd and Silver 1955; Thorstensson et al 1985) and in subjects with low back pain extensor muscle activity is increased near full flexion (Nouwen, Van Akkerveeken and Versloot 1987; Ahern, Hannon, Gorenzy, Follick and Parziale 1990). Forward trunk flexion with the hands reaching only as far as the knees would avoid the critical point and ensure a consistent muscle activity through this range.
Fast flexion, as distinct from slow flexion is initiated and controlled by anterior muscles especially the rectus abdominus (Thorstensson et al 1985).

The pelvis rotates forward over the legs with trunk flexion (Mayer, Tencer, Kristiferson and Mooney 1984; Waddell, Somerville, Henderson and Newton 1992; Porter and Wilkinson 1997) contributing as little as 33% (Mayer et al 1984) and as much as 50% (Gracovetsky, Newman, Pawlowsky, Lanzo, Davey and Robinson 1995) to trunk motion.
Extending the trunk from full flexion back up to upright stance proceeds as the opposite to forward bending though it involves greater muscle activity (Ortengren and Andersson 1977; Thorstensson et al 1985). Upon return to erect standing when lifting a weight, initial efforts begin with the hip extensors not the erector spinae (Floyd and Silver 1955).

Erector spinae muscles initiate extension from upright standing but the motion is then controlled by the abdominals acting eccentrically and synchronously with gravity (Ortengren and Andersson 1977; Thorstensson et al 1985; Crenna et al 1987). The hamstrings and calves are also involved in the initiation of trunk extension (Pedotti et al 1989). Return from extension requires a concentric contraction of the abdominals (Thorstensson et al 1985; Crenna et al 1987). During extension of the trunk, the pelvis also extends or rotates posteriorly at the hip joints (Mayer et al 1984; Waddell et al 1992).

The erector spinae muscles control lateral flexion, sometimes both acting simultaneously (Morris et al 1962) but principally by acting contralaterally (Floyd and Silver 1955; Thorstensson et al 1985). During fast side bending movements both the abdominals and extensors can be activated together. When performed slowly only the contralateral extensors work (Thorstensson et al 1985). Quadratus lumborum and intertransversarii are deep muscles inaccessible to surface EMG investigation; however their origin and insertion suggest a direct role in lateral flexion (Bogduk 1994). Using fine wire electrodes Andersson et al (1996) found contralateral quadratus activity during trunk side bending and ipsilateral activity while performing a lateral pelvic tilt in sitting.
In summary, muscle lines of action and the number and arrangement of bones upon which they act influence the outcome of movement and hence the resultant kinematics. Major trunk muscles attach anteriorly and posteriorly to the pelvis and thorax and not the lumbar spine and not to any vertebrae. The skeletal geometry of the trunk (pelvis, 12 vertebrae and rib cage) and the attachment and action of muscles along its length will determine trunk mobility and control. Major muscles attaching to the pelvis and ribs of the thorax may be a factor in central nervous system (CNS) control of movement and the kinematics of movement in normal trunk function and dysfunction. Hence it is possible that not only lumbar spinal kinematics and motor control but also the mobility and control of the thorax and pelvis together and separately may be a factor in the dispersion of forces within the trunk and may influence pain syndromes.

2.2 Measurement of Spinal Motion

Measuring spinal motion to determine trunk kinematics can be completed in vitro or in vivo using a variety of radiographic, hand held or electronic devices. Methods of measuring spinal motion and the devices employed in this study are presented prior to the discussion of trunk kinematics.

2.2.1 Methods for measuring spinal motion

Though fraught with lack of standards and inconsistent findings, the measurement of spinal range of motion is a major determinant for dysfunction in low back pain problems (Portek et al 1983; Pearcy 1986; Lowery, Horn, Boden and Wiesel 1992).
Studies of spinal motion have been undertaken with cadavers and the living using goniometric and radiographic measurements. Each method provides valuable information but each has limitations. Bogduk and Twomey (1991) point out that cadaver studies are limited by the removal of musculature and therefore do not accurately reflect motion in the living. Clinical studies are limited by the method of measurement and by the difficulty of reliably identifying bony landmarks (Bogduk and Twomey 1991), though markers placed on the skin have been shown to be consistent with radiographic recordings (Pearcy, Gill, Whittle and Johnson 1987). X-rays are accurate but such unnecessary exposure to radiation may be unethical.
Portek et al (1983) compared radiographic (biplanar radiography, radiography, vector stereography) and clinical techniques for measuring lumbar spinal motion (inclinometer, skin distraction, plumb line method) and reported wide discrepancies. Summarizing their study of six different spinal measuring techniques, these authors described most techniques as liable to large errors and as giving only indices of vertebral motion rather than true spinal movement. Use of the inclinometer was reproducible but required careful monitoring. Similarly, Mayer et al (1997) extensively defined and investigated sources of error with spinal measurement.

Four major sources for error were identified:

Spondylometer measurements (Twomey and Taylor 1983) may give lower ranges than radiographic techniques (Pearcy 1985). Electromagnetic devices in some cases provided exaggerated ranges as reported by Pearcy and Hindle (1989) and Hindle et al (1990) Buchalter et al (1989) and Nelson et al (1995) found lumbar flexion range of motion comparable to radiographic results. The method of attachment of sensors may account for these differences (Dolan and Adams 1993).
Measuring spinal motion can vary between each attempt. It has been suggested that test movements be practiced twice by each subject prior to measurement to ensure reliable and stable results (Keeley, Mayer, Cox, Gatchel, Smith and Mooney 1986). Also, spinal range of motion fluctuates throughout the 24 hour cycle. Testing performed between the hours of midday and 5 pm minimizes any altered flexibility due to circadian variation (Adams, Dolan, Hutton and Porter 1990; Russell, Weld, Pearcy, Hogg and Unsworth 1992).

2.2.2 The 3Space Tracker® and Motion Star®

The 3Space Tracker® and the Motion Star®are electromagnetic devices for measuring the position and orientation of a sensor with respect to a source in space (An, Jacobsen, Berglund And Chao 1988). The Tracker? has been shown to be an accurate, reliable and non-invasive clinical measurement device for spinal motion (Pearcy and Hindle 1989; McGill, Cholewicki and Peach 1997) and has been utilised in several spinal motion studies (Buchalter et al 1989; Pearcy and Hindle 1989; Hindle et al 1990; Pearcy et al 1993). Developed by Polhemus Navigation Sciences Division, McDonnell Douglas Electronics Company, the 3Space Tracker? provides three dimensional data of spinal motion (Pearcy and Hindle 1989). The device consists of an electronics unit which contains hardware that drives the system and software for the control of data collection, a source module and two sensors. The electronics unit determines the position and orientation of the sensors in the magnetic field produced by the source throughout an entire range of motion. Characteristic movement patterns are produced for the primary movement and accompanying planes (Pearcy and Hindle 1989) allowing movement progression and pattern to be studied as well as the end point of a movement.
A computer controls the operation of the 3Space Tracker, collecting, storing and analysing the 3D kinematic data (Pearcy 1993).
Similarly, the Motion Star (developed by Ascension Technology Corporation, Burlington VT, USA) is a later model electromagnetic measuring device that also determines the position and orientation of its sensors in 3D space. It captures the motions of up to 90 sensors simultaneously over long range without metallic distortion. Each sensor can be tracked up to 120 times per second.
Although reliably identifying bony landmarks can be a problem (Bogduk and Twomey 1991), the head, sacrum, manubrium and anterior superior iliac spine are large clear bony surfaces that are easily palpable. L1, T12 and T10 can be located by first defining the position of L4 at the level of the iliac crests and then counting upwards over the four spinous processes of L3-T12 (Burton 1986). C7 is the highest of the larger spinous process at the cerivo-thoracic junction.
Sensors have been placed at L1 (Pearcy and Hindle 1989; Hindle et al 1990) or T12 (Buchalter et al 1989; Waddell et al 1992; Porter and Wilkinson 1997) or T12/L1 (Nelson et al 1995) as representative of the upper limit of the lumbar spine.
A common method of application of the 3Space Tracker sensors has been with a strap that secures around the circumference of the trunk (Pearcy and Hindle 1989; Hindle et al 1990; Porter and Wilkinson 1997). However, some authors have reported errors with this method and designed alternative methods using adhesive tape to minimize these errors (Dolan and Adams 1993).
Sampling rate for use of the 3Space Tracker to measure spinal motion has ranged from 10 to 28Hz for 10 or 20 seconds (Table 2.1). The most frequent sampling rate and time, 10Hz for 10 seconds was chosen for study 1. For study 2 the sampling rate of 86Hz for 10 seconds was filtered to 5Hz and the time reduced to 6 seconds for one complete lifting cycle.

 

Table 2.1 Sampling rates for 3Space Isotrak (NA: Sampling time was not indicated in the report)

	Hindle  et al 1990	Pearcy & Hindle 1989	Buchalter et al 1989	Dolan & Adams 1993	Nelson et al 1995	Porter & Wilkinson 1997	Gatton & Pearcy 1999
Sampling Rate Hz	10	10	15	28	15	10	20
Sample time Seconds	10	10	NA	NA	NA	10	6

 

2.3 Kinematics of the Trunk

The following discussion of trunk kinematics begins with a description of the kinematics of individual trunk segments (pelvis, lumbar spine and thorax) followed by an outline of global kinematics and segmental relationships. Finally a brief review of trunk motor control is presented.

2.3.2 Pelvic kinematics

Rotation of the pelvis about the hips is a significant functional motion in human trunk action. In rising from a chair, the first motion is rotation of the pelvis and trunk about the femur (Schenkman 1990). Up to 50% of trunk flexion and extension in standing involves pelvic rotation (Mayer et al 1984; Gracovetsky et al 1995). Four to eight degrees of pelvic rotation in the sagittal plane occurs during human locomotion (Thurston and Harris 1983; Stokes, Andersson and Forssberg 1989). Sagittal plane pelvic rotation in sitting significantly effects cervical posture (Black, McClure and Polansky 1996). In the frontal and sagittal plane, pelvic rotation initiates balance control during sitting perturbations (Forssberg and Hirschfield 1994). Backward pelvic rotation initiates trunk extension from forward flexed position (Nelson, Walmsley and Stevenson 1995; McClure et al 1997). Bohannon, Gajdosik and LeVeau (1985) and Goeken and Hof (1991) reported an average of 32 and 25 degrees respectively for posterior pelvic rotation during a supine passive straight leg raising test. Pelvic rotation during trunk flexion has been shown to be altered in subjects with CLBP (Mayer et al 1984). Pelvic rotation in the sagittal plane deserves further and more complete investigation of its role in trunk kinematics in normal human function and altered function.
The term “pelvic tilt” has several meanings according to position, function and direction. In standing posture, pelvic tilt is a description of the static anterior-posterior position of the pelvis in the sagittal plane (Twomey and Taylor 1994). It has also been defined as the static position of the pelvis in the frontal plane (Waddell et al 1992).
Posterior pelvic rotation performed in the sagittal plane in crook lying and then held as a static position is a common reference for pelvic tilt (Kendall and McCreary 1983; Cailliet 1995). Both anterior and posterior pelvic rotation in the sagittal plane in this same position is a common position in which pelvic movements are assessed and utilized in the rehabilitation setting (Feldenkrais 1972; Kendall and McCreary 1983; Jull and Richardson 1994; Cailliet 1995). Rotation of the pelvis anteriorly and posteriorly can be performed on all fours (Jull and Richardson 1994), in standing (Cailliet 1995), or sitting (Oliver 1994) and still referred to as pelvic tilt. It can be performed in many different positions and variations for motor learning and awareness (Feldenkrais 1972, 1984).
In this study pelvic tilt was defined as anterior and posterior pelvic rotation in the sagittal plane in supine with bent legs.
Pelvic tilt in supine, though a regular rehabilitative exercise for low back pain, has not yet been the subject of a kinematic study nor is it measured clinically even though pelvic rotation has been shown to be diminished in subjects with CLBP (Mayer et al 1984). It is possible, that limited pelvic rotation in standing trunk movement may also be limited in supine. Pelvic tilt in supine is potentially a simple and easy position to observe and measure pelvic rotation in the sagittal plane. Rotating the pelvis on a firm horizontal surface provides consistent constrained conditions by which to measure pelvic tilt. Visual observation or hand held goniometric measurement of pelvic rotation in this position has the potential for a simple determination of pelvic mechanics otherwise difficult to determine in other functional positions and without sophisticated electronic equipment.
During trunk flexion, the pelvis rotates forwards about the femur during trunk flexion. Mayer et al (1984) using hand held inclinometers reported 63 degrees pelvic rotation during forward sagittal trunk motion. Similar findings have been reported by Waddell et al (1992), Esola, McClure, Fitzgerald and Siegler (1996) and Porter and Wilkinson (1997) (Table 2.2).

 

Table 2.2 Pelvic range of motion during sagittal plane trunk bending (in degrees). (f = female, m = male)

	Inclino-meter		Optoelectric device	3Space Tracker
Mean	Mayer et al 1984	Waddell et al 1992	Esola et al 1996	Porter & Wilkinson 1997
Flexion	63f&m	57f&m	70	58m
Extension	18f&m	-	-	-

During trunk extension the pelvis rotates backwards with the lumbar spine an average of 180 (Mayer et al 1984) (Table 2.1). Gracovetsky et al (1995) reported 4 degrees of pelvic rotation in the frontal plane accompanying 25 degrees of lumbar motion with trunk side bending.

2.3.3 Kinematics of the lumbar spine

Sagittal and frontal plane lumbar spine displacement and angular kinematics in standing are presented including ranges of motion. Using passive limits to trunk range of motion in the sagittal (experimenter applied force) and frontal planes (3 kg weight in hand) Dvorak, Panjabi, Chang, Theiler and Grob (1991a) measured displacement kinematics of the lumbar spine. For a normal or no low back pain (NLBP) population, total mean sagittal plane displacement (L1-2 to L5-S) was 54mm. In the frontal plane, displacement (L1-L2 to L4-L5) was 31.4mm (total left and right).
Flexion of the lumbar spine proceeds as an anterior vertebral rotation and an anterior translation in the sagittal plane (Kessen, During, Beeker, Goudfroois and Crowe 1984; Pearcy, Portek and Shepherd 1984). Flexion may proceed with all vertebrae moving together or a top down or bottom up or middle last sequencing (Gatton and Pearcy 1999). At the limit of flexion, the lumbar spine assumes a straight or slightly curved forwards position (Pearcy 1985) and hence the vertebral bodies become parallel to each other (Bogduk and Twomey 1991) (Figure 2.3).
Different measurement methods have produced variable data for lumbar flexion as summarized in Table 2.2. Pearcy et al (1984), using biplanar radiography, measured flexion range as 8-13 degrees for each level in 11 subjects with a total average flexion range of 51 degrees. Clinical measurements using inclinometers have determined a wide range of mean lumbar flexion from 22 to 55 degrees (Mayer et al 1984; Mellin 1990). Goniometric recordings for lumbar flexion from Mayer et al 1984 were between the reported electro-magnetic device and radiographic results whilst Mellin's (1990) goniometric recordings results seem extremely low. Mellin (1990) measured 52 females and males with a mean age 21.4 ± 1.6 years while Mayer's subjects were aged 19-51 with a mean age of 31 years. Hence age and subject were unlikely to affect these large differences in lumbar flexion range. Perhaps skill of inclinometer use may be a factor (Portek, Pearcy, Reader and Mowat 1983). Electronic goniometric measurement produces spinal ranges of motion very similar and sometimes a little larger than radiographic results (Pearcy 1985; Buchalter, Parpianpour, Viola, Nordin and Kahanovitz 1988; Pearcy and Hindle 1989; Nelson et al 1995).

Figure 2.3 Lumbar spine kinematics Schematic representation of the lumbar spine kinematics in the sagittal plane. A. Lumbar extension, E (to left), neutral stance (shaded) and flexion, F (to right). B. Anterior vertebral rotation of T12 during lumbar flexion. C. Anterior vertebral translation of T12 during lumbar flexion. Lumbar range of translation and rotation during trunk bending is obtained by subtracting the translation and rotation for T12 from the translation and rotation for the pelvis (Figure adapted from Pearcy 1985).

 

Table 2.2 Lumbar flexion range of motion (in degrees)

	3Space Isotrak			Biplanar radio-graphy		Inclino-meter
	Hindle et al 1990	Pearcy & Hindle 1989	Buchalter et al 1988	Pearcy 1985	Dvorak et al 1991a		Mayer et al 1984	Mellin 1990
Mean Flexion	67 f 74m	76	56	51	80		55	22 f 27 m

Extension of the lumbar spine is not as frequently performed in daily life activities as flexion and therefore perhaps has not been the subject of as much research (Crenna et al 1987). Extension has been stated to occur as the converse of flexion (Bogduk and Twomey 1991; McClure et al 1997) ie with posterior vertebral rotation and posterior translation.
McClure et al (1997) investigated extension of the lumbar spine from a flexed position and described the return to upright stance from full flexion as the reverse of flexion, however Nelson et al (1995) found pelvic rotation to initiate trunk extension while the lumbar spine actually flexed initially.
The range of extension has been measured as consistently less than flexion (Mayer et al 1984; Pearcy et al 1984; Pearcy 1985; Buchalter et al 1989; Pearcy and Hindle 1989; Hindle, Pearcy, Cross and Miller 1990). Only Mellin (1990) recorded extension as nearly twice that of flexion of other studies (Table 2.3) which may perhaps be explained by the large variability found when using inclinometers (Portek et al 1983).

 

Table 2.3 Lumbar extension range of motion (in degrees)

	3Space Isotrak			Biplanar radio-graphy	Inclino-meter
	Buchalter et al 1988 N=60 33f  27m	Pearcy & Hindle 1989 N=10m	Hindle  et al 1990 N=80 40f  40m	Pearcy 1985	Mayer et al 1984 N=13 6f  7m	Mellin 1990 N=103 48f  55m
Mean Flexion	22	23	24f   21m	16	27	40f  44m

Taylor and Twomey (1980) measured large numbers of 437 living subjects using a spondylometer. They did not differentiate between flexion and extension, instead reporting a mean combined sagittal rotation score of 42 degrees while Dvorak et al (1991a) measured a mean 80 degrees total sagittal plane motion using radiographic techniques. Dvorak et al (1991a) measured a passive (experimenter force at end of range) limit to flexion and extension. Buchalter et al (1989), Pearcy and Hindle (1989) and Hindle et al (1990) reported 78 degrees, 99 degrees and 95 degrees respectively for total active trunk sagittal plane bending. Hence the radiographic measurements of Dvorak et al (1991a) and the electromagnetic devices are comparable, however, Taylor and Twomey's (1980) lumbar sagittal motion is extremely low by comparison. This difference illustrates the errors inherent in the inclinometer method of spinal motion measurement (Portek et al 1983).
With lateral flexion of the lumbar spine, the upper vertebra rotates and translates in the frontal plane in the direction of motion (White and Panjabi 1978). As with all other lumbar motions, lateral flexion also eludes accurate determination by clinical measuring devices and can only be said to range between 35 and 60 degrees (Table 2.4). Radiographic techniques again result in low values for lumbar motion (Pearcy and Tibrewal 1984), while clinical measures vary from 42 degrees using an inclinometer (Mellin 1990) to 62 degrees utilizing the 3Space Isotrak (Hindle et al 1990) and automated video (Vachalathiti, Crosbie and Smith 1995). Though some accompanying lumbar rotation occurs with lumbar lateral flexion (Hindle et al 1990), the primary motion only will be investigated in this study.

 

Table 2.4 Lumbar spine frontal plane range of motion (in degrees)

	Radio-Graphy		3Space Isotrak			Inclino-meter	Auto-mated video
Authors	Pearcy & Tibrewal 1984	Dvorak et al 1991a	Hindle et al 1990	Pearcy & Hindle 1989	Buchalter et al 1988	Mellin 1990	Vachalathiti et al 1995
	N=10 10m	N=41 18f 23m	N=80 40f  40m	N=10 10m	N=60 33f  27m	N=103 48f  55m	N=100 54f  46m
Mean Lateral Flexion	35m	58	62f  58m	56m	47	42f  45 m	62

 

2.3.4 Kinematics of the thorax

The thorax as the composite of 24 ribs, a sternum and 12 thoracic vertebrae can be considered as a separate segment of skeletal motion with a range of movement (White 1969). Over 70 articulations and 204 possible combinations for movement ensure some mobility in the thorax (Anderson 1982). Based on experimental evidence, mathematical models and clinical experience, Lee (1993) proposed a biomechanical model for the thorax illustrated in Figure 2.3. This model is utilised here to describe sagittal and frontal plane displacement and angular kinematics of the thorax.

Figure 2.4 Biomechanical model for the thorax. Sagittal plane thorax rotation. A. Neutral position of the ribs and sternum in standing. B. Extension of the thorax is accompanied by a posterior rotation and elevation of the ribs and presumably the sternum. C. Flexion proceeds with anterior rib rotation and sternal depression (adapted from Lee 1993).

Occurring with a sagittal rotation and translation motion of thoracic vertebrae (Panjabi and White 1978), thorax flexion and extension proceeds with accompanying rib movements (Lee 1993).
As shown in Figure 2.4, theoretical extension of the thorax is coupled with a posterior rotation of the ribs (Lee 1993). Concomitantly, the sternum would follow this action and elevate and rotate backwards with extension. Conversely, with flexion, the ribs rotate anteriorly producing a depression of the sternum. This theoretical model for thorax motion is presumably valid in both supine and standing motion in the sagittal plane.

 

Table 2.5 Thorax sagittal and frontal plane range of motion (in degrees) * summated values from authors results. (f = females, m = males).

Device	In vitro	Literature Review	Inclino-meter		3Space Isotrak
Study	White 1969	White & Panjabi 1978	O'Gorman & Jull  1987 (in sitting)	Mellin 1990	Buchalter et al 1989
Sample size			N=120w	N=103 48w 55m	N=60 33w 27m
Flexion	20.4*	-	22	60m  60f	20
Extension	13.6*	-	20	4m -2f	20
Combined flexion-extension	34	76	42*	58-60*	40
Lateral flexion	-	72*	25f	70m 64f	40f&m

Total thorax sagittal range of motion has been recorded in vitro as 34 degrees (White 1969) with a flexion to extension ratio of 3:2. Panjabi and White (1978), however, from a literature review quote a total 76 degrees for sagittal thorax motion. They reported 4 degrees in each of the upper inter-vertebral levels, 6 degrees in the middle levels and 12 degrees in each of the lower levels (Table 2.5). In contrast, Buchalter et al (1989) found a total sagittal range of motion of 40 degrees equally distributed between flexion and extension. Mellin (1990) reported 60 degrees for flexion and only minimal extension of 4 to -2 degrees.

Motion of the thorax in the frontal plane results in frontal rotation of the thoracic vertebra and is theoretically accompanied by approximation of rib margins on the same side and rib separation on the opposite side. It is possible that the approximation of rib margins limits the motion of lateral flexion prior to vertebral limitations (Lee 1993). White and Panjabi (1978) quote the range of motion of lateral flexion to be fairly consistent throughout the length of this spinal segment at a mean of 6 degrees, though increasing slightly at lower levels. Buchalter et al (1989) found an overall average of 40 degrees in a group of 20 to 50 year olds. In sitting, O'Gorman and Jull (1987) reported 37 degrees lateral flexion in the 22 to 29 age group, while Mellin (1990) recorded 70 degrees for males and 64 degrees for females.
During lateral flexion, the thoracic vertebral spinous processes also rotate in the horizontal plane towards the concavity of the curve (Gregersen and Lucas 1967; White 1969). While lateral flexion may be accompanied by rotation, the primary movement of rotation in the frontal plane is much greater and will be considered in this study without its coupled motion.

2.3.5 Global kinematics of trunk movements

Lumbar, thoracic and pelvic motion is most often measured individually and few studies have investigated the motion of these segments simultaneously. Presented below are some studies that have measured trunk segment motion simultaneously. These studies include measurement of pelvic and lumbar motion during trunk flexion (lumbopelvic rhythm or lumbopelvic ratio) and extension and thorax motion recorded simultaneously during trunk motion.
The ratio of lumbar to pelvic contributions to trunk flexion has been investigated (lumbar range / pelvic range = L-P ratio). Mayer et al (1984) reported lumbopelvic ratio (L-P ratio) over two periods of trunk flexion: up to 90 degrees flexion and 90-120 degrees and found L-P ratios of 1.72 and 0.17 respectively (Table 2.6). Lumbar movement was greater than pelvic motion during the first 90 degrees of trunk flexion and then pelvic motion predominated in the later ranges of trunk flexion. These results have been repeated in other studies using more sophisticated electronic equipment to find similar L-P ratios (Esola et al 1996; Porter and Wilkinson 1997; Granata and Sanford 2000) (Table 2.6). Authors vary in what reference point to use for the lumbar spine. The upper limit for the lumbar spine has been selected as L1, T12-L1, T12, T10 and C7 which may produce varying results depending on thorax contributions to trunk bending (Table 2.6). Presumably, in choosing T10 or C7 for the upper limit for the lumbar spine authors are making the assumption that the thorax does not contribute to trunk bending. This may not be the case. Gracovetsky et al (1995) reported a continuous constant equal L-P ratio throughout trunk flexion and actually calculated a trunk-pelvic ratio rather than an L-P ratio (Table 2.6). Similarly, Granata and Sanford (2000) in determining L-P ratio defined the lumbar spine as L5 to T10. This may be different to the actual true L-P ratio. Nelson et al (1995) reported that at 50% of full trunk flexion, the lumbar spine had flexed to 94% of its maximum and the pelvis had reached 84% of its maximum. These figures are difficult to compare to other studies using direct ratios between lumbar and pelvic motion.
Farfan (1975) suggested that pelvic and lumbar movement occurred sequentially while Nelson et al (1995) reported lumbo-pelvic rhythm to be more sequential during lifting (extension) and more simultaneous during lowering (flexion) a 9.5kg box to 90% to full trunk flexion. Pelvic backward rotation always preceded lumbar extension during lifting a 9.5kg load (Nelson et al 1995). Granata and Sanford (2000) found simultaneous motion for these segments during trunk bending.
Load may alter lumbo-pelvic mechanics. Granata and Sanford (2000) reported the lumbar spine to contribute 70 percent of forward bending (pelvic 30%) which increased when weight was added to the task. This contrasts with Gracovetsky et al (1995) who found no affect on spinal coordination with loading.
Few studies have reported simultaneous pelvic, lumbar and thoracic motion. One spinal motion study did actually record cervical, thoracic and lumbar motion simultaneously for standing flexion, extension, rotation and lateral flexion using an electromagnetic sensing device (Buchalter et al 1989). All spinal segments, cervical, thoracic and lumbar contributed to each trunk movement. Interestingly, only for flexion was lumbar range of motion considerably more than thoracic motion. For extension and lateral flexion, thoracic and lumbar motion was almost equal, and for rotation, thorax motion was more than four times that of lumbar motion (Buchalter et al 1989).
Tully and Stillman (1997) using computer aided video analysis studied two groups of subjects; one group that could touch their toes and one that could not. Lumbar and thorax movements were measured during a toe touching test. The authors determined five different patterns of thorax movement during trunk flexion. (Dividing the motion into three stages, initial, middle and final thirds of the test, patterns were defined as follows: flexion-extension-extension, flexion-extension-flexion, extension-extension-flexion, flexion-flexion-flexion and extension-flexion-flexion). The lumbar spine consistently flexed during the test. Interestingly, subjects who could successfully reach their toes utilized more thorax extension (mean –4.0 ± 14.7 degrees) whilst the unsuccessful toe touchers used more thorax flexion (mean = 4.0 ± 10.2 degrees and overall mean for both groups = 0.0 ± 12.9 degrees).

 

Table 2.6 Lumbo-pelvic ratio *Lumbar-pelvic ratio = lumbar motion/pelvic motion.

Measurement device	Inclinometer	Opto-electric mdevice			Tracker	Electro-magnetic device
Study	Mayer et al 1984	Gracovetsky 1995		Esola et al 1996	Porter & Wilkinson 1997	Granata & Sanford 2000
Upper limit for lumbar spine	T12-L1	C7		T12-L1	L1	T10
Range of trunk flexion	Lumbo-pelvic ratio*
0-30	-	1	2		-	2.4
30-60	-	1	1		-	2.4
60-90	-	1	0.5		-	1.8
50% flexion	-	1	-		-	-
Full flexion	-	1	-		-	-
0-90	1.7	1	-		2	-
90-120	0.17	1	-		0.33	-

 

It is difficult to determine adequate reasons for these discrepancies in flexion motion. Methods, pitfalls and developments in measuring spinal motion are discussed in section 2.5 Measurement of Spinal Motion. It is also possible that different tasks demand different contributions of thorax motion.

Cavanaugh, Shinberg, Ray, Shipp, Kuchibhatla and Schenkman (1999) investigated trunk kinematics during an asymmetrical functional reach task. Subjects were required to reach as far forward as possible without falling. Young subjects (N = 34, mean age = 28) reached significantly further than the elderly subjects (N = 32, mean age = 70 years). The elderly displayed significantly less thorax rotation and centre of mass (COM) displacement.
Many authors have previously reported the thorax to contribute little to spinal movement (Davis et al 1965; Munro 1965; Grieve 1988; Cailliet 1995; Lindh 1989) and it is usually modelled as a rigid segment (Gracovetsky and Farfan 1986). This assumption of thorax rigidity may be justifiable when considering the minimal motion found by studies like Tully and Stillman (1997) but not when considering the 20-60 degrees reported by Buchalter et al (1989) and Mellin (1990). What seems possible is that under some circumstances of trunk movement the thorax is required to move minimally and under others it moves considerably more depending on the task and method of performing that task. This was observed by Cavanaugh et al (1999) during the previous described functional reach task. Pre-test trunk range of motion measurements were taken for the thoracic and lumbar spines. During the maximal reach task, only 30 and 50 percent of available rotation and lateral flexion respectively were utilised.
Recently, reports on spinal modelling and stability have acknowledged the role of the long erector spinae muscles attaching along the length of the rib cage (Bergmark 1989; McGill 1992; Kiefer, Shirazi-Adl and Parnianpour 1998). It seems possible that if long muscles of the trunk attach from the pelvis to each rib that some leverage in muscle action may be gained from a flexible rib cage. However in these studies the thorax is modelled as a rigid segment and the question of thoracic mobility and leverage was not explored. One possible reason for modelling the thorax as a rigid segment is to reduce the complexity of the analysis in favour of a simplistic workable mathematical model.

2.3.6 Pelvic and thorax motion during pelvic tilt

Jull and Richardson (1994) comment that with a posterior pelvic tilt, the anterior rib cage depressed downwards. Abdominal muscles contracting to tilt the pelvis, may affect the position of the sternum and ribs pulling them downwards. Similarly, motion of the pelvis may be transmitted through the lumbar and thoracic vertebrae to determine sternal and thorax kinematics (Figure 2.5).
Activation of the long erector spinae muscles to produce anterior pelvic tilt and hence extension of the lumbar spine may also extend the thorax and tilt the sternum upwards. Black et al (1996) reported a relationship in sitting between pelvic motion and resultant cervical spine posture but did not determine the intermediary thorax kinematics.
Feldenkrais (1972) described a corresponding matched motion of the pelvis, thorax and head with pelvic anterior and posterior tilting. Feldenkrais proposed that coordination between the head and pelvis via the length of the spine and chest is essential for efficient strain free spinal movement. Simultaneous supine pelvic rotation and corresponding thorax motion has not been previously investigated.

Figure 2.5 Pelvic tilt and the thorax. Schematic representation of pelvic tilt. Posterior and anterior pelvic tilt produces both displacement and angular kinematics. A. Posterior pelvic tilt pushes through the vertebrae to produce a cranial translation and a depressive rotation of the sternum. B. Anterior pelvic tilt pulls on the vertebrae to produce a caudal translation and a cranial rotation of the sternum.

 

2.3.7 Trunk kinematics of a functional lifting task

Lifting a box from several different platforms (floor, knee, hip and chest height) has been a frequent method of studying the biomechanics of the spine for nearly 50 years (Floyd and Silver 1955; Nussbaum et al 2000). Whilst trunk kinematics during lifting has been studied extensively, investigation of trunk segmental interactions is sparse. For example, Tsuang, Schipplein, Trafimow and Andersson (1992) modelled body segments during a lifting task. Markers on the hand, elbow, greater trochanter, humerus, knee and ankle provided data for a five bar linkage model. However, no segmental trunk information was allowed for in this methodology. Similarly, Khalaf et al (1999) developed a methodology for evaluating lift characteristics, using markers on the ankle, knee, hip, shoulder and wrist. A model was formulated from these markers and hence again the trunk was modelled as one rigid segment.
Nussbaum et al (2000) measured “torso” kinematics during a lifting task using markers on L5/S1, T10 and T4. The “torso” was defined as T4 and T10 relative to the pelvis; however only one score for torso kinematics was presented without identifying what segment was considered to be the torso.
Interactions between voluntary and postural kinematics during perturbed lifting has been investigated in an asymptomatic population (Oddsson, Persson, Cresswell and Thorstensson 1999). Subjects were required to lift a 20kg box from floor to desk and were perturbed in the anterior and posterior directions. The study found large erratic erector spinae muscles activity during a backward perturbation. Oddsson et al (1999) proposed that this rapid switch from voluntary contractions to postural reactions as a possible mechanism of injury to the low back. Again, inter-segmental trunk kinematics was not determined in this study.
Gracovetsky, Kary, Pitchen, Levy and Ben Said (1989) found increased EMG activity in the spinal muscles and increased compressive stress within the spine, when the lumbosacral lordosis of the spine increased or decreased beyond an optimal level during a lifting task. The model used predicted that for every angle of flexion, there was a unique amount of pelvic tilt to minimize compressive forces within the lumbar spine.
Several studies of trunk kinematics during lifting tasks have measured lumbar spine motion as being between T10 and the pelvis and sometimes the trunk as the motion from the pelvis to T10 (Marras, Lavender, Leurgans, Rajulu, Allread, Fathallah and Ferguson 1993; Fathallah et al 1998; Granata and Sanford 2000). In these studies the lumbar spine is modelled with three extra vertebrae and the trunk is missing 10 thoracic vertebrae. If all studies consistently use this method then perhaps this protocol is a valid one. Many other studies, however, have used either T12 (Buchalter et al 1989; Waddell et al 1992; Porter and Wilkinson 1997) or L1 (Pearcy and Hindle 1989; Hindle et al 1990; Dolan and Adams 1998) or both (ie between T12 and L1) (Mayer et al 1984; Nelson et al 1995; Esola et al 1996) to represent the upper limit of the lumbar spine. In lifting and balance studies, the trunk is most often defined as being between C7 or the shoulder and S2 or the greater trochanter (Anderson 1982; Allum, Bloem, Carpenter, Hullinger and Hadders-Algra 1998; Diener et al 1990; Alexander et al 1992; Tsuang et al 1992; Khalaf et al 1999; Oddsson et al 1999; Rietdyk, Patla, Winter, Ishac and Little 1999). Defining trunk segments and assumptions about movement contributions of the thorax could influence results. To save confusion, inconsistent or incorrect results, it may be better to maintain a consistent anatomically correct definition for kinematic studies of lumbar and trunk motion. The current study defined the trunk as that segment between C7 and the pelvis and the lumbar spine as between the pelvis (S2) and T12 (Study 1) or L1 (Study 2).

2.3.8 Motor control of the trunk

Trunk action from standing involves a coordinated sequencing of most of the body segments: the head, trunk, pelvis and legs (Crenna et al 1987; Pedotti et al 1989; Oddsson 1990). Babinski (1899) first reported the accompaniment of hip and knee motion with trunk bending. In the sagittal plane trunk bending proceeds as a coordinated action of the head and trunk in the direction of the intended movement and the pelvis and knees in the opposite direction (Figure 2.6). The displacement of the pelvis and legs opposite to the intended movement of the trunk occurs to maintain postural equilibrium (Crenna et al 1987). With voluntary action in standing, muscle activation and body segment kinematics occur to achieve the intended command but also to control posture and balance. Similarly, Belenki, Gurfinkel and Palistev (1967) showed that raising an arm was preceded by a backward movement of the trunk and legs aimed at minimizing the disturbance of balance due to the movement. The accompaniment of trunk bending with these adjustments in posture is called “anticipatory postural adjustments” (Massion 1992). The intended movement may also cause reaction forces and inter-segmental interactions through the posturo-kinematic chain resulting in associated movement separate from the voluntary intention (Massion, Alexandrov and Vernazza 1998).
It has been shown that a burst of muscle activity in the lower leg, thigh and trunk precedes the onset of movement of that segment (Crenna et al 1989; Oddsson and Thorstensson 1986). Hence, because muscles act before movement rather than in response to movement, it is proposed that trunk axial synergies are centrally controlled and of the anticipatory postural adjustment type (Massion et al 1998).
Using a rigid three link model Ramos and Stark (1990) determined that the backward movement of the pelvis and legs during the forward motion of the trunk occurred as a necessary result of the mechanical relationships between these structures as well as the result of anticipatory postural adjustments for maintenance of postural stability.
The relationship between ankle, knee and hip angle has been shown to be linked by fixed ratios that could be reduced to one controlling parameter (Alexandrov, Frolov and Massion 1998). These authors suggested that only one degree of freedom was needed to control trunk bending and proposed two hypotheses for explaining this occurrence. This fixed ratio could result from passive biomechanical constraints or a central command adjusted to the biomechanical properties of the individual. That is to say that both neurological and mechanical factors may determine trunk kinematics.
Lumbar lordosis and trunk angle have been determined as strictly correlated suggesting that the CNS controls the degree of freedom of the spine (Mitnitski, Yahia, Newman, Gracovetsky and Feldman 1998). Mitniski et al (1998) measured lumbar and trunk angle while subjects lifted varying weights (bar 11-68kg) from 10cm height to mid thigh to find that lumbar lordosis and trunk angle were highly correlated and therefore limiting the degrees of freedom to be controlled by the CNS. Hence the results of Mitniski et al (1998), Alexandrov et al (1998) and Ramos and Stark (1990) suggests that CNS must control both trunk angular and displacement kinematics.
Trunk motion follows a bell shaped velocity curve (Oddsson 1988). During slow sagittal (one second for forward flexion) and frontal plane movements of the trunk, motion proceeds sequentially craniocaudally and during fast motions the sequencing is simultaneous (Thorstensson et al 1985; Crenna et al 1987). Hence speed of motion may affect motor control of the trunk.
During trunk flexion and lateral flexion, pelvic action, lumbar elongation and lordosis proceed in consistent patterns despite loading and range of movement (Gracovetsky et al 1995). Similarly, Vernazza, Alexandrov and Massion (1996) showed that the final position of the centre of mass (COM) during a trunk bending task remained constant despite changes in load applied to the shoulders. Hence subjects predicted the effect of additional load and adjusted the kinematics accordingly to maintain the COM within the support base.

Figure 2.6 Coordination of the trunk and pelvis during axial bending. With forward flexion, the pelvis and knees displace backwards, while the trunk and head translate forwards. While with extension, the pelvis and knees translate forwards and the head and trunk displace backwards. The eyes lead trunk motion.

Trunk motion can be performed consistently in similarly reproducible patterns. Six repeated trials with the same subject produced consistent motion patterns for the head, trunk, pelvis and legs during trunk extension (Oddsson 1988). Even over two days trunk bending in the sagittal plane has shown to be consistent as illustrated by coefficient of variance scores less than 0.20 for pelvic and lumbar in 10 subjects (Nelson et al 1995). These features of trunk motion suggest a well defined pattern.
The position of C7 relative to the pelvis in the sagittal and frontal planes in standing and lying down is well controlled by the nervous system. Standing and supine positioning of C7 relative to the pelvis in the frontal plane is highly consistently controlled (Jakobs, Miller and Schultz 1985).This study found standing subjects were able to return C7 to its neutral upright position to within 3.1mm or 0.3 degrees subtended at the sacrum after passive and active repositioning in standing and sitting in the frontal plane. In the horizontal plane subjects could rotate the trunk by rotating the shoulders relative to the hips or the hips relative to the shoulders to 2.8 ± 1.2 degrees and 2.9 ± 1.4 degrees respectively. Through each stride in human gait, the trunk remains vertical to within ±1.5 degrees in the sagittal and frontal planes (Winter and Eng 1995).
Trunk motion is known to be altered in some movement trained individuals (Pedotti et al 1989; Mouchnino, Aurenty, Massion and Pedotti 1993). Dancers maintain trunk alignment to the vertical during standing leg abduction to 45 degrees while untrained individuals tilt their trunk significantly with abduction of the leg (Mouchnino 1993). Gymnasts are able to perform trunk movements on a narrow beam without loss of balance while untrained individuals cannot (Pedotti et al 1989). Muscle synergies during backward trunk bending and control of COM in trained subjects showed more highly flexible and adaptable patterns than the untrained.

2.4 The Effect of Age and Gender on Trunk Kinematics

Generally, range of spinal motion decreases significantly with age (Taylor and Twomey 1980; Twomey and Taylor 1983; O'Gorman and Jull 1987; Hindle et al 1990; Gracovetsky et al 1995; Vachalathiti et al 1995) (Tables 2.7 and 2.8). There have been no reported gender differences for thorax motion (Buchalter et al 1989) while lumbar spine studies have reported less lumbar flexion in females (Wolf, Basmajian, Russe and Kutner 1979; Burton and Tilloston 1988; Hindle et al 1990; Vachalathiti et al 1995).

2.4.1 The pelvis

With advancing age, the pelvic contribution to forward bending increases, corresponding to a decrease in lumbar motion and a decrease in the ability to reduce lumbar lordosis. There is little change in pelvic motion over the fourth and fifth decades (Gracovetsky et al 1995).

2.4.2 The thorax

O'Gorman and Jull (1987) investigated the thoracic motion of 120 adult females in sitting. As illustrated in Table 2.7, flexion decreased significantly from the twenties to the thirties, remained consistent over the next two decades (30-50 years) and then dropped significantly beyond age 50. Extension decreased significantly from the third through the fifth decade but then altered minimally beyond the fifth decade. Buchalter et al (1989) found no significant relationship between age, gender and range over the third and fourth decade for thorax motion.

 

Table 2.7 Thorax mobility with age in females. (L = left, R = right) (Age in years, range in degrees, O'Gorman and Jull 1987).

Age (N= 120)
years              decade	22-29                     3rd	30-39                      4th	40-49                     5th	50-59                     6th
Flexion	33	25	23	16
Extension	37	29	21	20
Lateral flexion	37	L 30   R 33	26	22

 

2.4.3 The lumbar spine

Hindle et al (1990) investigated lumbar mobility and found a consistent loss of range of lateral flexion with age in both sexes, while extension was decreased in females only (Table 2.8). Flexion did not significantly reduce with age in either gender group. Vachalathiti et al (1995) studied three age groups (20 to 25, 36 to 59 and >60) and as illustrated in Table 2.8, flexion and lateral flexion did not significantly change with age until beyond 60 years. Again males had more flexion than females which has also been previously reported (Wolf et al 1979; Burton and Tilloston 1988; Hindle et al 1990).

Table 2.8 Lumbar mobility with age and gender *significance level p = 0.05, ** p = 0.01. †significant for > 60 compared to < 60 years ‡ significant for > 60 compared to >35 years.

study	Hindle et al 1990
	females			males
age range of motion	20-29	30-39	40-49	20-29	30-39	40-49
flexion	59**	70	64*	75	73	77
extension	32	24	20	26	17	24
lateral flexion	62	54	53*	58	53	47
study	Vachalathiti et al 1995
	females			males
age range of motion	20-35	36-59	>60	20-35	36-59	>60
flexion	39	37	32†	48	46	33 †
lateral flexion	66	62	49‡	67	62	53‡

 

2.5 Kinematics of Chronic Low Back Pain

2.5.1 Definition

The problem of persistent CLBP is so great that it eludes a clear pathophysiological definition and is currently only recognizable by its persistence despite investigation and treatment. For example, CLBP has been defined as back pain persisting for longer than three months (Waddell et al 1992), pain persisting for greater than 49 days (Quebec-Task-Force-on-Spinal-Disorders 1987) or two years work loss (Mellin 1988).
CLBP is also described as the end stage of spine problems (Mayer 1991) and in terms of disability and societal costs (Frymoyer and Cats-Baril 1991; Tulder et al 1995). It is most prevalent and persistent through the 4th and 5th decade of life (30 to 50 years) (Long, BenDebba and Torgenson 1996). A person with CLBP and off work for six months has a 50% chance of a return to work, while only a 10% chance after 12 months and no chance after two years (Beals and Hickman 1972). Hence once low back pain becomes established it becomes more difficult to ameliorate.
CLBP is a major issue for the individual, health professions and the community. Ongoing determination of contributing factors is necessary to better define and manage the problem. Therefore it is imperative that a better understanding of the contributing factors be developed. One development in the understanding of CLBP has been the identification of an altered kinematic relationship between pelvic and lumbar forward bending in subjects with CLBP (Mayer et al 1984; Dolan and Adams 1993; McClure et al 1997; Porter and Wilkinson 1997). It is possible that other contributory kinematic relationships exist between motion segments adjacent to the lumbar spine.
The skeletal segment immediately adjacent superiorly to the lumbar spine is the thorax. As described earlier, the powerful muscles of the trunk span over the lumbar spine from the pelvis to the thorax and hence all three segments coordinate in trunk function and may each contribute to altered spinal kinematics relationships in CLBP.

2.5.2 Clinical diagnostic tests

There are many clinical tests that successfully differentiate the low back pain sufferer from normal individuals, however as yet, none can differentiate low back pain from CLBP (Waddell, Allan and Newton 1991). The most common positive clinical tests for CLBP are reduced flexion, extension (pelvic plus lumbar rotation), lateral flexion, straight leg raising range of motion and sit up weakness (Waddell et al 1992). Hip joint immobility (Mellin 1988; Dolan and Adams 1993) and an altered lumbopelvic rhythm (Mayer et al 1984; Porter and Wilkinson 1997) have also been identified in CLBP sufferers.
Those physical parameters that differentiate low back pain from CLBP have not yet been defined. Possibly there may not be any physical characteristics that indicate that back pain will become chronic, persistent and unresponsive to treatment. Other parameters such as social and environmental features may be more involved in chronicity than physical characteristics.
One of the problems in identifying signifi