Symmetry of Timing of Hip and Lumbopelvic Rotation Motion in 2 Different Subgroups of People With Low Back Pain

Participants 

A convenience sample of 39 people (23 men, 16 women; mean age, 28.1±8.0y), with chronic or recurrent LBP,23 participated in the study. The sample was a subset of people with LBP who had been recruited from the community for a larger study. The primary focus of the larger study was on the relationship between participation in rotation-related activities and types of impairments displayed. All subjects were people who reported a history of LBP for a minimum of 1 year. Subjects also reported (1) regular participation (minimum 2 times a week) in a sport that placed repetitive rotational demands on the hip and lumbopelvic region, and (2) an increase in their LBP symptoms related to their sport that occurred either during or after play. Examples of activities that were considered to place rotational demands on the hip and lumbar region included sports such as tennis, squash, racquetball, and golf. No subject was in an acute flare-up23 of an LBP problem at the time of testing. A flare-up in the current report is defined as a phase of pain superimposed on a recurrent or chronic course which consists of a period, usually a week or less, when the LBP is markedly more severe than usual for the patient. Using a forced choice format, people were screened for participation through a telephone survey and then with a list of questions asked on the day of testing. We excluded people from participation if they reported that a physician had diagnosed them with any of the following spine-related conditions: (1) previous spinal surgery, (2) marked kyphosis or scoliosis, (3) spondylolisthesis, (4) spinal stenosis, (5) spinal instability, (6) spinal fracture, (7) ankylosing spondylitis, (8) degenerative disk disease, (9) disk herniation, or (10) serious spinal complications (eg, tumor or infection). People were also excluded if they had been diagnosed by a physician with (1) rheumatoid arthritis, (2) any condition resulting in severe neurologic involvement, (3) neurologic disease which required hospitalization, (4) history of unresolved cancer, (5) osteoporosis, or (6) were currently pregnant. Finally, people who reported any lower-extremity impairment, for example, impairment secondary to a prior surgery or leg-length discrepancy, were also excluded. Subjects read and signed an informed consent statement approved by the Washington University School of Medicine Human Studies Committee before participating in the study.

Procedures 

Self-report measures 

Subjects first completed 4 self-report measures. One measure was a numeric rating scale which consisted of an 11-point scale (range, 0–10) provided verbally to the person. The person rated his current symptoms in standing and his average and worst symptoms over the prior 7 days, with higher numbers indicating higher symptom intensity.24 A second measure was the Oswestry Disability Index (ODI),25 a measure of perceived LBP-related disability. The total score on the ODI can range from 0% to 100% (0% indicating no disability). A third measure was the Baecke Habitual Activity questionnaire,26 which is a 16-item questionnaire that measures habitual physical activity over a prolonged period of time. Finally, a fourth measure was a questionnaire of demographic and LBP history variables recommended for reports of studies involving people with LBP.27 Summary data related to self-report measures are presented in table 1.

Table 1. Differences in Characteristics of Subjects in 2 LBP Subgroups

	Variable	LBP Subgroup		Statistical and Probability Value
		Lumbar Rotation (n=13)	Lumbar Rotation With Extension (n=26)
	Age (y)	29.0±7.9	27.7±8.2	t37=−.47, P=.64
	Sex (male/female)	6/7	17/9	χ12 test=1.33, P=.25
	Body mass index (kg/m2)	25.5±5.5	24.4±2.5	t37=−.88, P=.39
	Right handed (%)	92	96	χ12 test=.26, P=.61
	Positive for magnified symptom behavior32 (%)	0	0	Statistical test not performed due to lack of variability in data.
	Positive for neurologic involvement (%)	0	0	Statistical test not performed due to lack of variability in data.
	Symptom intensity⁎ (0–10)
	Current in standing	2.3±1.9	2.3±1.6	t37=.10, P=.92
	Average over prior 7 days	3.6±2.0	3.0±1.7	t36=−1.00, P=.32
	Worst over prior 7 days	5.0±2.0	4.7±2.4	t36=−.48, P=.63
	Location of symptoms (%)
	Low back only†	100	100	Statistical test not performed due to lack of variability in data.
	Duration of LBP (y)	6.4±5.7	6.1±5.7	t36=−.16, P=.88
	Reproduction of LBP symptoms during hip lateral rotation test (%)			χ32 test=1.34, P=.72
	No increase in LBP	38	35
	Right HLR	23	11
	Left HLR	15	27
	Both	23	27
	Ability to modify increased symptoms with primary test of hip lateral rotation test during secondary test‡ in clinical examination (%)	100	100	Statistical test not performed due to lack of variability in data.
	Episodes of LBP23 (prior 12mo)	14.4±22.0	6.2±8.5	t36=−1.70, P=.10
	ODI scores25 (0%–100%)	21.2±17.7	13.4±6.1	t36=−2.03, P=.05§
	Median	18.0	12.0
	Range	64.0	26.0
	Frequency of participation in rotation-related sport (sessions per week)	3.0±1.9	3.0±1.6	t37=.00, P=1.00
	Duration of participation in rotation-related sport (min/session)	92.3±25.9	100.4±60.3	t37=.46, P=.65
	Baecke activity score26 (3–15)	8.2±0.7	8.4±0.9	t37=.58, P=.57
	Movement time of HLR angle (s), mean ± standard error	Left = 2.1±0.3	Left = 2.6±0.2	F1=.01, P=.93 (subgroup by side)
		Right = 2.1±0.3	Right = 2.6±0.2

NOTE. Values are mean ± standard deviation (SD) or as otherwise indicated.

⁎ Verbal numeric rating of symptom intensity on a scale of 0 (no symptoms present) to 10 (worst imaginable symptoms).24 † Low back region is defined as the area between T12-L1 interspace and the gluteal fold.52 ‡ Secondary tests are conducted if the patient reports an increase in symptoms with a primary test. Modifications for movement tests involve restricting or eliminating lumbar region movement during a trunk or limb movement while encouraging movement in other segments, for example, the hips or thoracic region. Modifications are accomplished using verbal cues, trunk muscle activation by the person, and manual assistance by the examiner. The person reports his symptoms with each secondary test relative to the symptomatic primary test. § Indicates a significant difference between the 2 LBP subgroups.

Laboratory testing 

Subjects then participated in laboratory testing. We used a 3-dimensional, 6-camera, motion measurement systema to examine the kinematics during the HLR test. The motion measurement system has a sampling rate of 60Hz for each camera and a static resolution of 1mm for a volume of 1m3. Movement of the lower extremity and lumbopelvic region were captured during the HLR test through the use of 20 retro-reflective markers placed on pre-determined anatomic landmarks of the trunk, pelvis and extremities (appendix 1, fig 1).

View Large Image Download to PowerPoint

Fig 1. Kinematic model indicating angles for HLR and lumbopelvic rotation. Abbreviations: α, HLR angle (in degrees); GT, greater trochanter marker; IC, iliac crest marker; K, knee marker; LM, lateral malleolus marker; θ, lumbopelvic rotation angle (in degrees). From Gombatto et al.53 Reprinted with permission.

A series of laboratory-based movement tests were conducted in the larger study, including the HLR test. The HLR test was performed in prone. The subject’s lower extremity was passively positioned in 90° of knee flexion and neutral hip abduction and adduction and neutral rotation. Subjects were asked to laterally rotate their hip so to bring their foot in toward the midline as far as possible, and then return it to the starting position. Hip movements were performed at a self-selected speed. A 10-second period was allotted to complete the hip movement and the movement was performed once with each extremity. No subject exceeded the 10 second limit. Reports of symptoms with the HLR test were obtained. Symptoms with the HLR test were relative to the subject’s symptoms in prone. The possible options for symptom reports included the following: (1) remained the same, (2) increased, (3) decreased, or (4) eliminated.

Laboratory Data Processing 

Initially, all marker data were filtered using a fourth-order, dual-pass, Butterworth filter with a cutoff frequency of 2.5Hz. The initial cutoff frequency was chosen because the movements being measured were relatively slow. Because subjects performed the test movement at self-selected speeds, a filtering frequency based on individual movement times was then used. Based on the HLR movement time, raw data were filtered at a subject-specific cutoff frequency28 (fcss) that was calculated by taking the reciprocal of 15% of the period:

Clinical Examination and Classification 

Subjects then participated in a standardized clinical examination used to classify the person’s LBP problem.18, 19 Testing of the reliability of examiners performing the physical tests and measures from the standardized examination has been reported.19 Fair to good reliability29 of classifying a patient’s LBP has also been reported (κ=.57; percentage of agreement, 78%)30 by our own research group, and an independent research group (κ=.56).31

The examination includes tests of movements and positions, neurologic screening, and tests to screen for magnified symptom behavior.32 Each test is proposed to be associated with a particular direction of lumbar region movement or alignment, including flexion, extension, rotation, or a combination of rotation and flexion or rotation and extension. Because rotation and lateral bending are coupled motions in the lumbar spine,33, 34 symptoms or patterns of movement or alignment identified during examination tests associated with either rotation or lateral bending currently are labeled rotation. Symptoms are monitored with each test, and for some of the tests, judgments of patterns of movement and alignment are made. The judgments are made by the examiner based on visual or on both visual and tactile information. The response options for LBP symptoms are the same as those described for the laboratory testing. With each initial test, referred to as a primary test, the subject assumes a position or performs a movement using a preferred strategy. If the subject reports an increase in symptoms with the primary test, it is immediately followed by a secondary test in which the subject’s preferred strategy is standardly modified.35, 36 The secondary test is performed in an attempt to decrease or eliminate the subject’s symptoms. Based on the consistency of findings of lumbar region movement and alignment and LBP symptom behavior across the clinical examination, each subject is classified into 1 of 5 LBP subgroups.18 The subgroups are named for the directions of movement and alignment that appear to contribute to the LBP problem. The subgroups include (1) Rot, (2) lumbar flexion, (3) lumbar extension, (4) RotFlex, and (5) RotExt. In deciding a person’s classification, priority is given to the symptomatic primary tests that can be successfully modified (symptoms decreased or eliminated). Findings typical of the Rot and RotExt LBP subgroups are described in table 2.

Table 2. Typical Findings of People in the Lumbar Rotation and Lumbar Rotation With Extension LBP Subgroups

	Responses With Tests	Lumbar Rotation Findings	Lumbar Rotation With Extension Findings
	Symptom reports⁎	•Symptoms are increased with 1 or more primary test† associated with each one of the following directions of lumbar region alignment or movement: (1) flexion, (2) extension, and (3) rotation. •Symptoms are decreased or eliminated with secondary tests‡ of symptom-provoking primary tests associated with lumbar region flexion, extension or rotation. •Symptoms are often decreased or eliminated with positions that decrease compression of the lumbar region (eg, performing a push-up in sitting; assuming hook lying; assuming quadruped).	•Symptoms are increased with 1 or more primary tests associated with each one of the following directions of lumbar region alignment or movement: (1) extension and (2) rotation. •Symptoms are decreased or eliminated with secondary tests of symptom-provoking primary tests associated with lumbar region extension and rotation. •Symptoms typically remain the same with primary tests associated with lumbar region flexion. •Increases in symptoms with primary tests of limb movements are typically different on the right versus left (asymmetric symptom behavior).
	Judgments of patterns of alignment and movement	•Lumbar region rotation patterns are evident with various primary tests across the examination and with symptom-provoking functional activities.	•Lumbar region rotation and extension patterns are evident with various primary tests across the examination and with symptom-provoking functional activities. •Lumbar region rotation patterns are often asymmetric.

⁎ In deciding a person’s LBP classification, priority is given to the directions of movement and alignment associated with symptomatic primary tests that can be successfully modified (symptoms decreased or eliminated) during the associated secondary test. † During primary tests the person performs either a trunk or limb movement, or assumes a position using his preferred strategy and reports his symptoms relative to a designated reference movement or position. The movement tests from the examination are active movements unless otherwise specified. ‡ Secondary tests are conducted if the patient reports an increase in symptoms with a primary test. Modifications involve either positioning the lumbar region in as close to a neutral alignment as possible, or restricting or eliminating lumbar region movement during a trunk or limb movement while encouraging movement in other segments, for example, the hips or thoracic region. Modifications are accomplished using verbal cues, trunk muscle activation by the person, and manual assistance by the examiner. The person reports his symptoms with each secondary test relative to the symptomatic primary test.

Kinematic Measures 

Hip and lumbopelvic rotation 

We calculated hip lateral rotation and lumbopelvic rotation angles across each 10-second trial of HLR. Hip lateral rotation was calculated by tracking the lower leg segment across time. The lower leg segment was defined by a vector from a marker on the lateral aspect of the knee joint line to a marker at the distal aspect of the lateral malleolus. Angle α was calculated as the change in angle of the lower leg segment across time, relative to its initial position (see fig 1). Assuming that no relative motion occurs between the tibia and femur, movement of the tibia should reflect rotation of the femur. Degree of HLR was then defined by subtracting, from angle α, pelvic motion that occurred in the plane of the lower leg. Thus, movement of the pelvis that resulted in movement of the lower leg was not included in the hip rotation calculation.

Lumbopelvic rotation was calculated by tracking the pelvic segment across time. The pelvic segment was defined as a vector formed by the right and left iliac crests. Relative to the initial position of the pelvic segment, the position of the pelvic vector was calculated across time during the HLR motion to calculate θ, the degree of lumbopelvic rotation (see fig 1).

The criteria that we used to identify the optimal start and end of movement for hip and lumbopelvic rotation were determined through an iterative process. The hip lateral rotation angle and lumbopelvic rotation angle for each subject was plotted against time. Start and end points for each angle were identified using threshold criteria of (1) angular displacement and (2) angular velocity. The threshold values were varied and the start and end of both angles, for each subject, were plotted and visually inspected. For visual inspection, an accurate start point was defined as the time point at which there was a consistent change in the slope of the angle-time plot. An accurate end point was defined as the first time point at which the angle-time plot appeared to reach the absolute maximum. The threshold values for angular displacement and velocity that resulted in the most accurate start and end points for the majority of subjects were used in the final algorithm. The start of movement for HLR was defined as the time at which both (1) angular displacement of the lower leg segment exceeded a threshold of 1° and (2) angular velocity exceeded 5% of the maximum angular velocity for the lower leg segment. The start of movement for lumbopelvic rotation was defined as the time at which both (1) the angular displacement of the pelvic segment exceeded a threshold of 1° and (2) the angular velocity exceeded 15% of the maximum angular velocity for the pelvis segment. The end of movement for the HLR and lumbopelvic rotation angles were defined by the first point at which each angle reached 99% of its maximum during the hip rotation movement.

We tested the reliability of HLR and lumbopelvic rotation measures in a sample of 10 subjects without a history of LBP. The intraclass correlation coefficient model 3,1 (ICC3,1)37 and standard error of the measure38, 39 were used to index reliability. The values for each motion, for each extremity, were found acceptable and are reported in table 3.

Table 3. ICC37 Values and Standard Error Measures38, 39 and for Right and Left HLR and Lumbopelvic Rotation

	Kinematic Measure	ICC3,1	Standard Error of Measure (deg)
	Left HLR angle	.99	1.5
	Right HLR angle	.82	0.8
	Left lumbopelvic rotation angle	.97	1.4
	Right lumbopelvic rotation angle	.80	0.8

From Gombatto et al.53 Reprinted with permission.

Data Analyses 

Dependent variables 

Select descriptive variables related to subject characteristics, self-report, examination, and laboratory measures were examined. Maximum HLR and lumbopelvic rotation angle were calculated. The difference between the start time of HLR and the start time of lumbopelvic rotation (startdiff) was calculated (fig 2). Because each individual moved at a self-selected speed, to compare the timing of lumbopelvic rotation relative to HLR across subjects, the startdiff variable was normalized to HLR movement time (startdiff/HLR movement time) (see fig 2).

View Large Image Download to PowerPoint

Fig 2. Angle of HLR and lumbopelvic rotation as a function of time. Abbreviations: LPR, lumbopelvic rotation; startdiff, difference in time between the start of HLR and lumbopelvic rotation.

Kinematic Variables 

We compared the 2 subgroups with regard to maximum HLR and lumbopelvic rotation angle. The maximum angles for each subgroup, for each side, are provided in table 4. There were no differences between subgroups for maximum HLR angle (Rot, 41.9°±1.8°; RotExt, 45.1°±1.3°; F1=2.08, P=.16), and there were no differences between sides (left, 43.7°±1.2°; right, 44.3°±1.2°; F1=.03, P=.86). There was also no interaction of subgroup and side for maximum HLR angle (see table 4). There also were no differences between subgroups for maximum lumbopelvic rotation angle (Rot, 4.8°±0.8°; RotExt, 6.0°±0.6°; F1=1.42, P=.24), and there were no differences between sides (left, 5.3°±0.5°; right, 5.9°±0.6°; F1=.98, P=.33). There was no interaction of subgroup and side for maximum lumbopelvic rotation angle (see table 4).

Table 4. Differences of HLR and Lumbopelvic Rotation During Kinematic Testing Between People in the Lumbar Rotation LBP Subgroup Versus People in the Lumbar Rotation With Extension LBP Subgroup

	Variable	LBP Subgroup				Statistic
		Lumbar Rotation (n=13)		Lumbar Rotation With Extension (n=26)
		Left	Right	Left	Right
	Active HLR (deg)	42.3±2.0	41.5±2.1	44.5±1.5	45.8±1.5	F1=.73, P=.40 subgroup by side⁎
	Lumbopelvic rotation (deg)	4.8±0.8	4.9±1.0	5.6±0.6	6.5±0.7	F1=.55, P=.46 subgroup by side⁎

NOTE. Values are mean ± SD.

⁎ Indicates the statistical results for the 2-way interaction of subgroup by side. There were also no main effects of LBP subgroup or side.

We then examined the correlation coefficients indexing the relationship between right and left startdiff values for each subgroup and the coefficient of determination for each subgroup. People in the Rot subgroup showed more symmetry in timing of HLR and lumbopelvic rotation, right versus left, than people in the RotExt subgroup. The correlation coefficient indexing the relationship of right and left startdiff for the Rot subgroup was .94 and the r2 value was .88 (P=.00) (fig 3A). The correlation coefficient for the RotExt subgroup, on the other hand, was .31 and the r2 value was .10 (P=.12) (fig 3B). The correlation for the Rot subgroup differed significantly from the correlation for the RotExt subgroup (z=3.74, P<.01).

View Large Image Download to PowerPoint

Fig 3. Difference in start time of lumbopelvic rotation relative to start time of HLR, startdiff, during right versus left HLR for people in 2 LBP subgroups. (A) Relationship between right and left startdiff* values for people in the lumbar rotation LBP subgroup. (B) Relationship between right and left startdiff* values for people in the lumbar rotation with extension LBP subgroup. *Startdiff value was normalized to HLR movement time.

Model for Development of LBP 

The MSI classification system for LBP is based on the kinesiopathologic model of movement (KPM),40 a conceptual framework for the mechanisms proposed to contribute to the development of musculoskeletal pain. The KPM, therefore, provides a theoretical basis for treatment of musculoskeletal pain problems, including LBP. The basic premise of the KPM is that musculoskeletal pain develops as a result of loss of movement precision, that is, an alteration in appropriate biomechanics, at 1 or more segments in a specific anatomic region. Loss of movement precision is proposed to develop when movements are repeated and postures are sustained in a particular direction(s) during everyday activities, inducing changes in neural and musculoskeletal components of the movement system. Such changes then result in adoption of direction-specific strategies of movement and alignment that become generalized across many activities. The specific loss of precision proposed for development of LBP is an increase in flexibility of 1 or more lumbar segments in a specific direction(s), with a potential decrease in flexibility in other regions, for example the thoracic region or hips. The low magnitude loading with repetition in the same direction is considered to contribute to acceleration of accumulation of tissue stress in the lumbar region due to minimal time without loading,41 microtrauma, and eventually LBP symptoms. The model further contends that until the specific precision loss is addressed, the symptoms have the potential to persist or recur. The subgroups proposed in the MSI classification system display the most prevalent directions of movement and alignment strategies that are proposed to characterize LBP problems.

Treatment Based on KPM 

Based on the KPM, identifying a person’s LBP subgroup provides a basis for treatment. Specifically, to improve a person’s LBP problem, treatment would be directed at correcting the direction-related movements and alignments that characterize the person’s particular LBP subgroup. The RotExt subgroup has been described as characterized by a tendency to move and align the lumbar region in rotation and extension across several examination tests and with functional activities. The rotation-related movements and alignments are also often asymmetric18 (see table 2). In general, asymmetry has been proposed to be a potential mechanism for LBP because of the increased frequency of loading of one or more lumbar segments on one side compared with the other.42 In the current study the RotExt subgroup displayed an asymmetry in timing of hip and lumbopelvic region rotation, right to left, with the HLR test. In particular, lumbopelvic rotation occurred early in the range of hip lateral rotation in 1 lower extremity compared with the other (see fig 3B). Specific to the HLR test findings, the precision loss is a relative increase in lumbopelvic rotation flexibility compared with hip rotation flexibility evidenced by lumbopelvic rotation early in the range of HLR motion with 1 lower extremity compared with the other. Treatment would be directed at the neural and musculoskeletal components that appear to contribute to the asymmetry with the HLR test, as well as with other examination tests and functional activities. The Rot subgroup, on the other hand, has been described as characterized by a tendency to move and align the lumbar region in rotation across several examination tests and with functional activities18 (see table 2). The potential mechanism for LBP is the increased frequency of loading lumbar region segments in rotation due to adoption of rotation-related strategies across many activities. The Rot subgroup in the current study displayed symmetry in the timing of hip and lumbopelvic rotation, with the majority of subjects displaying lumbopelvic rotation early in the range of hip lateral rotation motion with both the right and left lower extremities (see fig 3A). Specific to the HLR test findings, the specific precision loss is the increase in lumbopelvic region rotation flexibility relative to hip rotation flexibility bilaterally for the majority of subjects. Treatment would be directed at the neural and musculoskeletal components that appear to contribute to the repeated lumbopelvic region rotation with the HLR test, as well as similar patterns of lumbopelvic rotation with other examination tests and functional activities.

ODI Scores 

The Rot subgroup had a slightly higher mean ODI score (P=.05) (see table 1) than the RotExt subgroup. The difference in LBP-related disability could pose an alternative explanation for the differences in symmetry of timing of hip and lumbopelvic region rotation between the 2 subgroups. The difference in ODI scores, however, was found to be the result of one case in the Rot subgroup that was identified as an outlierb (ODI score, 68%). Similar to other subjects in the Rot subgroup, however, the timing of lumbopelvic and HLR movement, right versus left, for the outlier case was similar (right startdiff, .22s; left startdiff, .20s). Thus, it does not appear that a difference in LBP-related disability is the reason for the differences between the 2 LBP subgroups in symmetry of timing.

Prior Studies 

Some earlier studies have examined the relationship between active hip lateral rotation and LBP.2, 4, 5 The variable examined in each study, however, has been the amount of end range hip lateral rotation motion and not the timing of movement of the hip and lumbopelvic region during hip lateral rotation. For example, Mellin2 reported no relationship between the amount of end range active hip lateral rotation and reported baseline symptom levels in people with chronic or recurrent LBP. Mellin4 also reported no differences in amount of end range active hip lateral rotation between young adults with and without a history of LBP. Chesworth et al,5 on the other hand, reported decreased end range active hip lateral rotation in people with LBP compared with matched subjects without a history of LBP. The current study differs from these prior studies in that (1) both active hip lateral rotation and lumbopelvic rotation were measured, (2) different subgroups of people with LBP were compared, and (3) relative timing of hip lateral rotation and lumbopelvic rotation between sides was the variable of interest rather than end range hip lateral rotation. Because of these differences, the current findings regarding active hip lateral rotation cannot be directly compared with prior studies. In the current study, however, we focused on the relative timing of active hip lateral rotation and lumbopelvic rotation between sides for a number of reasons. First, our clinical observation18 and prior data suggest that a number of people with LBP report an increase in LBP with the active HLR test.43 Second, we have reported that some people with LBP also display early lumbopelvic rotation with HLR.44, 45, 46 Third, the lumbopelvic rotation that occurs with active HLR appears to be of importance because restricting lumbopelvic rotation during the HLR test results in a decrease or elimination in symptoms in the majority of people with LBP.35, 36 Fourth, the prevalence of lumbopelvic rotation during active hip lateral rotation appears to be related to the types of activities in which people with LBP participate regularly.47 Specifically, it was found that a larger proportion of people with LBP who participated in asymmetric activities showed asymmetry of lumbopelvic rotation with the HLR test than did people with LBP who participated in symmetric activities.

Some of the prior studies have also examined the relationship between end range hip medial rotation and LBP. The findings from these studies suggests that in various LBP subgroups, decreased passive or active end range hip medial rotation,2, 5, 9 or a decrease in end-range passive medial rotation compared with lateral rotation in 1 extremity may be related to the LBP problem.6, 10 Finally, a discrepancy in the amount of passive hip medial rotation between sides has been associated with a particular type of LBP that appears to respond positively to spinal manipulation.13, 14 We have documented that people with LBP report an increase in LBP symptoms with the test of active hip medial rotation.36, 43, 48 Similar to active hip lateral rotation, we have observed that a factor that appears to contribute to the increase in symptoms with the active hip medial rotation test is the attempt to move the lumbopelvic region during the hip motion.36, 45 At the time of the current study, however, we did not have data regarding the pattern of lumbopelvic rotation with hip internal rotation. Future studies could focus on quantifying, with clinical and instrumented measures, the amount and timing of lumbopelvic motion that occurs with active hip medial rotation in people with LBP, and its relationship to LBP symptoms.

Symmetry of Movement 

Trunk lateral bending in standing is the test movement in which symmetry of trunk movement is most often examined.16, 18, 20, 21, 49, 50 Asymmetry of trunk lateral bending is considered to be an important finding in determining the classification of a person’s LBP problem.16, 18, 49 In contrast to prior work, we have examined symmetry of trunk movement, indexed by lumbopelvic movement, with an active limb movement test. The current data suggests that a limb movement, HLR, can induce movement of the lumbopelvic region, and that people in the RotExt subgroup display more asymmetry of lumbopelvic rotation than people in the Rot subgroup, at least with hip lateral rotation. Future studies could examine (1) the symmetry of lumbopelvic movement with other limb movement tests in people in the Rot and RotExt LBP subgroups, (2) the relationship of symmetry with trunk movement tests to symmetry with various limb movement tests, and (3) the relationship of symmetry with movement tests to symmetry with more functional movements such as gait. Findings from such studies could provide additional information on which to base treatment of people in the Rot and RotExt LBP subgroups.

Prevalence of LBP Subgroups 

The LBP subgroups of interest in the current study are only 2 of 5 subgroups described in the MSI classification system for LBP. The prevalence of the different LBP subgroups has been proposed to differ, with the majority of people described as having an extension, rotation, or combined rotation and extension LBP problem.18 Two prior studies provide partial support for the described distribution of subgroups. A factor analysis of examination data from 188 people with LBP resulted in the largest proportion of variance explained by tests related to the RotExt subgroup, followed by tests related to the Rot and the Ext subgroups.46 A subsequent study of 51 people with LBP examined on their first physical therapy clinic visit resulted in the following distribution of LBP subgroups: (1) RotExt (41%), (2) RotFlex (22%), (3) lumbar extension (18%), (4) Rot (14%), and (5) lumbar flexion (5%).36 In the sample (N=43) considered for the current study, 13 (30%) of the subjects were classified as Rot, 26 (61%) were classified as RotExt, and 4 (9%) were classified as RotFlex. No subjects were classified into the lumbar flexion or extension subgroup. Thus, the distribution of the current sample is similar to our prior studies. We did not include the RotFlex subgroup in the current analyses, however, because (1) our hypothesis was focused specifically on the Rot and RotExt subgroup findings and (2) of the limited number of people in the RotFlex subgroup.

Study Limitations 

The current study has some potential limitations. First, we examined the symmetry of hip and lumbopelvic movement during a standardized clinical movement test. Whether these findings generalize to functional activities is not known at this time. Future studies could explore the relationship between movement characteristics identified during standardized movement tests and movement characteristics with functional activities. Second, the findings from our study may not generalize to all people with LBP. The people in the current study were people with LBP recruited from the community who regularly participated in a rotation-related sport. Our findings, therefore, may not be applicable to all people referred to physical therapy clinics or to people who do not regularly participate in rotation-related activities. We do know from prior studies, however, that some subjects recruited from clinics who reported participation in a variety of activities could be classified into the Rot or RotExt LBP subgroups.30, 44 Finally, it is possible that the movement patterns we have identified do not contribute to the person’s LBP and may instead be the result of having LBP symptoms. If this were true, intervention directed at the movement patterns would likely not impact the person’s LBP problem. We have found, however, that when movement patterns that characterize an LBP subgroup are modified during the clinical examination and with treatment, symptoms improve for the majority of people with LBP.35, 36, 51 We also know from data obtained with the clinical examination in prior studies,35, 36 as well as the current study, that all subjects in both LBP subgroups who reported an increase in symptoms with HLR reported a decrease or elimination of symptoms when their movement pattern during HLR was modified to restrict lumbopelvic rotation (see table 1). Such findings suggest that how subjects were moving during the HLR test appears to be relevant to their LBP symptoms.