Archives of Physical Medicine and Rehabilitation
Volume 87, Issue 9 , Pages 1170-1176, September 2006

Are Lumbar Repositioning Errors Larger Among Patients With Chronic Low Back Pain Compared With Asymptomatic Subjects?

  • Malin Åsell, MS

      Affiliations

    • Centre for Musculoskeletal Research, University of Gävle, Gävle, Sweden
    • Svenska Rygginstitutet AB, Sundsvall, Sweden
    • Department of Odontology, Clinical Oral Physiology, Umeå University, Umeå, Sweden.
    • Corresponding Author InformationReprint requests to Malin Åsell, MS, Svenska Rygginstitutet AB, Box 7010, SE-850 07, Sundsvall, Sweden
    • ,
  • Per Sjölander, PhD

      Affiliations

    • Centre for Musculoskeletal Research, University of Gävle, Gävle, Sweden
    • Southern Lapland Research Department, Vilhelmina, Sweden
    • ,
  • Helmut Kerschbaumer, PT

      Affiliations

    • Svenska Rygginstitutet AB, Sundsvall, Sweden
    • ,
  • Mats Djupsjöbacka, PhD

      Affiliations

    • Centre for Musculoskeletal Research, University of Gävle, Gävle, Sweden

Article Outline

Abstract 

Åsell M, Sjölander P, Kershbaumer H, Djupsjöbacka M. Are lumbar repositioning errors larger among patients with chronic low back pain compared with asymptomatic subjects?

Objective

To resolve the debate over whether lumbar repositioning acuity is reduced in patients with chronic low back pain (CLBP) by using a study design and methodology to minimize the effects of potential confounders.

Design

A single-blinded, controlled, multigroup comparative study.

Setting

Vocational rehabilitation center.

Participants

Ninety-two patients with CLBP, divided into subgroups based on severity of symptoms and diagnostic characteristics. An age- and sex-matched group (n=31) of healthy subjects were the control.

Interventions

Not applicable.

Main Outcome Measures

We measured repositioning errors (variable, constant) at 3 positions of the lumbar spine. Subjects were guided to a sitting target posture and asked to perform lumbar flexion before reproducing the target posture. Self-assessed pain, self-efficacy, and functional ability were addressed through questionnaires.

Results

There were no differences in repositioning errors between the patients with CLBP or the subgroups of patients and the control group. We found only weak correlations between the repositioning errors and the self-reported data on functional disability, self-efficacy, and pain.

Conclusions

We suggest that sensorimotor dysfunctions in CLBP should be evaluated with methods other than repositioning tests in order to generate data relevant to the development of rational diagnostic methods and rehabilitation programs.

Key Words:  Back pain , Low back pain , Lumbosacral region , Proprioception , Rehabilitation

 

SEVERAL REPORTS SUPPORT that low back pain (LBP) is associated with sensorimotor disturbances that involve altered muscle coordination and proprioceptive deficits of the trunk. To enable rational assessment and rehabilitation of musculoskeletal dysfunctions, different methods for objective measurements of various sensorimotor parameters have been adopted. It is generally believed that measurement of repositioning acuity provides an estimate of the precision of the proprioceptive system. There is strong evidence that repositioning acuity of the cervical spine is disturbed in chronic neck pain.1, 2, 3, 4 It is controversial, however, whether the repositioning acuity of the lumbar spine is altered in chronic low back pain (CLBP). Some studies have reported significant differences between patients with CLBP and healthy subjects,5, 6, 7, 8 while others have failed to show any differences.9, 10, 11

There are probably several reasons for this discrepancy, but small and heterogeneous samples of patients were used in several studies.5, 6, 7, 9, 10 There were also considerable methodologic differences between the studies. Lumbar repositioning tests were performed in sitting,7, 9 in 4-point kneeling,5 and in standing with and without immobilization of the lower extremities.5, 10, 11 There are reasons to believe that cutaneous cues due to equipment, clothes, and hair may have biased the results,5, 7, 9, 10, 11 particularly because a recent study has found that subtle cutaneous inputs may have major effects on postural control.12

In previous studies, the starting and the target positions of the repositioning tests were not standardized with respect to the neutral sitting posture and the maximum range of motion (ROM) of the lumbar spine. Although repositioning errors are influenced by movement distance, no study of patients with CLBP has controlled for this confounder.

Lumbar repositioning acuity has mostly been assessed by measuring the absolute error.5, 6, 8, 9, 10 Because the absolute error is calculated as the average of the absolute values of the matching errors for the trials, it contains a mix of information on both the variability in the matching errors over the trials, as well as systematic deviations from the target. Repositioning variability (variable error) is usually calculated as the standard deviation (SD) of the matching errors, while the systematic deviation from the target (constant error) is calculated as the mean deviation from the target.13 There is evidence that the constant error reflects systematic biases in the motor control system, while the variable error reflects noise (acuity) of the position sense.14 Also, according to psychophysical research, the variable error represents the sensory discrimination threshold in adjustment tasks.15 Thus, in comparison to the absolute error, the variable error is a more valid measure of proprioceptive acuity. Only in 3 previous studies with patients with LBP has the variable error been analyzed; in 2 studies it was analyzed together with the absolute error11, 16 and in 1 it was analyzed along with the constant error.7

Classifications of LBP, reviewed by Riddle,17 entail several different approaches. From a clinical perspective, it would be useful if clinical research more often addressed subgroups of CLBP as defined by commonly used classification approaches. This could contribute to a better understanding of the characteristics of different subgroups. In Sweden in general, as well as at the vocational rehabilitation center where this study was performed, the McKenzie method18 is commonly used as an instrument for diagnoses and treatment. Another classification used to tailor an individual patient’s rehabilitation is the International Classification of Diseases, 10th Revision (ICD-10). Both the McKenzie method and the ICD have acceptable reliability.19, 20

Taken together, disparities in study design, selection of patients, and methods of error assessment are likely reasons for the reported different results on repositioning acuity in CLBP. Therefore, our main purpose in this study was to investigate whether patients with CLBP show larger lumbar repositioning errors compared with asymptomatic subjects, and to assess the clinical relevance of lumbar repositioning tests as measures of proprioceptive ability in such patients. Another objective was to reveal potential relationships between measures of repositioning error (constant error, variable error), clinical symptoms, and self-reported function. The study was designed to avoid methodologic shortcomings of previous investigations regarding the numbers of patients and trials, inclusion criteria, cutaneous cues, and standardization of the test procedure. We included a large sample of patients with well-defined CLBP to enable analyses of specific subgroups of patients.

Back to Article Outline

Methods 

The study was designed as a single blinded, controlled, multigroup comparative study.

Participants 

Ninety-two consecutive patients (table 1) referred to a vocational rehabilitation center (Rygginstitutet in Sundsvall, Sweden) for rehabilitation of CLBP were included in the study. To obtain an age- and sex-matched control group, 31 healthy subjects were recruited through announcements at different workplaces in Sundsvall (see table 1). The study was approved by the ethics committee of the medical faculty of the University of Umeå, and all subjects who met the inclusion criteria gave their informed consent to participate.

Table 1. Characteristics of the Study Sample
CharacteristicsControl (n=31)CLBP (n=92)
Men1645
Women1547
Age36±938±7
Lumbar ROM (deg)⁎38±1033±9‖
Repositioning angle (deg)†18.5±5.515.8±4.4‖
Time for repositioning (s)‡16.8±3.616.6±3.5
Pain intensity (mm VAS)§ 54±24
ODQ sum score 29±12
SES mean score 6.3±1.7
SF-36
Physical function99.2±2.969.8±22.2
Role–physical93.8±24.634.5±43.5
Bodily pain89.7±16.848.9±29.0
General health85.8±12.667.5±19.8
Vitality75.6±16.650.6±23.6
Social functioning89.8±21.474.9±24.9
Role–emotional83.3±32.859.8±43.0
Mental health81.6±19.671.9±19.4

Abbreviations: ODQ, Oswestry Disability Questionnaire; SES, Self-Efficacy Scale; SF-36, Medical Outcomes Study 36-Item Short-Form Health Survey; VAS, visual analog scale.

NOTE. Values are n or mean ± standard deviation.

⁎ Range of maximal lumbar motion in flexion and extension.

† The distance between the turning and the target position.

‡ The movement time for the repositioning, that is, the time from leaving the target position to indication of returning to the target position.

§ Assessed on a blank VAS ranging from 0 (no pain) to 100mm (worst possible pain).

‖ P<.01.

The inclusion criteria for the CLBP group were: major complaints of pain and discomfort in the lower back region, with chronic or recurrent episodes of pain lasting more than 6 months. The inclusion criteria for the control group were: healthy, with no complaints of back pain in the last year, as assessed with the Nordic Council of Ministers questionnaire.21

Exclusion criteria for both groups were diabetes, psychiatric diagnoses, neurologic and rheumatic disorders, dizziness, and vestibular diseases; subjects who had been surgically treated within the last 3 months were also excluded. This information was obtained from medical records, interviews, and questionnaires.

Diagnoses and Characteristics 

The patients were classified into subgroups according to their symptoms and impairments19 and to the McKenzie evaluation outcomes.18, 20 We used the medical diagnoses from the last physician who had examined the patient before he/she was referred to the rehabilitation center. The diagnoses were based on status indices as described by Riddle17 and on the patient chart notes from the physician, in most cases complemented with ICD-10 codes. These included lumbago (n=42; ICD-10 codes M54.5, M54.9); lumbago sciatica (n=21; ICD-10 codes M54.3, M54.4); disk-related pain (n=22; ICD-10 codes M51.3, M53.9), and spondylolisthesis (n=7; ICD-10 code M43.1), verified by radiology according to the physicians’ medical charting.

Nonspecific diagnoses of CLBP, to which lumbago can be referred, are most common, with a prevalence of about 90% to 95% in the early phase of the chronic condition.22 In our sample of patients, the proportion of nonspecific medical diagnoses was lower (44%), which seems plausible considering that they had experienced LBP for a considerable time before being referred to the vocational rehabilitation center. Lumbago sciatica, disk-related pain, and spondylolisthesis have objective impairments in common and are therefore more appropriately referred to as specific medical diagnoses.

All patients were classified using the McKenzie18 method by specially educated physiotherapist. According to this classification, most patients had either derangements (n=38) or dysfunctions (n=27). The postural syndrome was identified in 1 patient only, while 27% of the patients’ symptoms were characterized as inconclusive.

The Repositioning Test 

All subjects underwent the same test for assessment of the lumbar repositioning acuity. This was measured with a sitting pelvic-tilt test, where the degrees of freedom are reduced compared with when the repositioning test is performed in a standing posture, hence promoting more isolated and well-defined lower back movements. In a sitting posture, afferent input from joints and muscles in the limbs is minimized. The sitting posture is also a common body position and pelvic tilt is a familiar movement for most people; this implies that the test is easy both to comprehend and to perform. The subjects were seated on a wooden stool adjusted so that the hips and knees were positioned at 90° (fig 1A).

  • View full-size image.
  • Fig 1. 

    Schematic illustration of the repositioning test. The subjects were guided to (A) the target position, which was actively repositioned after (B) a close to maximal lumbar flexion (90% of maximum flexion). The boxes at the spinal cord represent the location of the Fastrak receivers (S2, T7, middle). (C) Maximal ROM (solid lines), the normal sitting position (S2nor), the target position (dashed line), and 90% of maximal flexion (dashed line).

The repositioning acuity was measured with a Fastraka electromagnetic tracking system, which has been shown to have good sensitivity and reliability in studies of cervical and lumbar motion.23, 24 Using the method of Hoppenfeldt,25 the physiotherapist identified and marked S2, T7, and the mid-point between those 2 segments. At these positions, Fastrak receivers were attached to the skin with double-sided self-adhesive tape (fig 1B).b

The normal sitting posture was recorded while subjects sat in their normal upright position. Subjects were then asked to stand up, and then to sit down. This procedure was repeated twice and the lumbar angles in the normal posture were calculated as the mean values (S2nor; fig 1C).

The lumbar ROM was measured at S2 as the angle between maximal flexion and extension (S2maxext, S2maxflex; fig 1C). Before the measurements were taken, all subjects were verbally instructed how to perform the movement and were manually guided by the physiotherapist during a test movement to ensure a pelvic-tilt movement without any major anteroposterior movement of the thoracic spine. The ROM was tested twice for each subject and an average value was calculated.

The target position was set one third of the way toward maximal extension (S2maxext) from the subjects’ normal sitting position (S2nor, see figs 1A, 1C). Hence, the target position (TPS2) was calculated according to:

To ensure that subjects maintained an upright position of the thoracic and cervical spine during the testing procedure, we set a reference position for the T7 receiver. This position was calculated as the mean position in the sagittal plane for the T7 receiver after 3 test trials (see below) in which the physiotherapist guided the subject, by hand, to the predetermined target posture (TPS2) while maintaining an upright position of the subject with the aid of an online display of the lumbar angle (S2).

The physiotherapist explained the procedures to the subject before the test. Subjects were given 6 practice trials, 3 with verbal instructions from the experimenter, and 3 with prerecorded instructions only. A single trial was performed according to the following: after the physiotherapist had guided the subject to the target position he/she was asked to memorize the position for 2 seconds (see fig 1A). Then the subject was instructed to flex the lumbar spine until an auditory signal was delivered, which occurred at 90% of the subject’s S2maxflex (see fig 1B). This signal indicated that the subject should return to the target position. The participant was instructed to match the target position as closely as possible. When the subject thought the target position was reached, he/she gave a verbal signal. This position was denoted as the matching position. In total, 10 such trials were performed by each subject, with a 3-second rest period between each trial. During each trial, data from the Fastrak system were continuously stored on a personal computer to enable off-line calculation of repositioning errors and movement times.

Subjects were blindfolded during the tests to exclude visual input. To minimize any biasing effect of potential inconsistencies in the instructions, they were prerecorded on tape and played automatically during the test. No verbal feedback was given during the test sessions. Cutaneous sensory input was minimized by having subjects wear only undergarments, by putting long hair in a ponytail or bun, by using bolstered rolls in the armpits to minimize contact from arms and hands, and by using minimal patches of double-sided, self-adhesive tape to fasten the receivers to the skin. Subjects were instructed not to eat or drink for 2 hours prior to testing to minimize input from mechanoreceptors in the abdominal region. The experimenter who guided subjects during the test sessions was blinded to their back pain status, as well as to their scores from the questionnaires.

Questionnaires 

Patients with CLBP made pain drawings to indicate the number of lumbar pain regions, the area with the most dominant pain, and the areas with continuous pain, intermittent pain, and/or itching pain (fig 2). Pain drawing is a valid method with which to evaluate the quality and spatial dimension of pain.26 We assessed pain intensity before and after the repositioning test with a blank 100-mm visual analog scale.

All participants were assessed with Medical Outcomes Study 36-Item Short-Form Health Survey,27 a questionnaire designed for self-assessment of health status and quality of life. It has 36 questions divided into 8 subscales, each with a score ranging from 0 to 100 (see table 1). Specific functional disabilities of the patients were evaluated by the Oswestry Disability Questionnaire (ODQ)28 and the Low Back Outcome Score (LBOS).29 The ODQ asks 10 questions to be answered on a 6-grade scale (range, 0–5) from which the invalidity grade is calculated (degree of individuality range, 0–100). The LBOS has 12 items divided into 2 sections. Each item is answered on a 4-grade scale, with higher scores indicating greater disability. Patients’ self-efficacy was assessed with the Self-Efficacy Scale (SES),30 a 20-item scale with responses ranging from 0 (not confident) to 10 (very confident). A mean score is calculated, with high scores indicating much confidence.

Data Management and Analysis 

The variable error and the constant error were used as outcome measures for the repositioning test. The algebraic error (AlgErr) for each trial was calculated as the difference between the target position and the position reached after the lumbar flexion, according to the formula:

where i represents a single trial, matching position (MP) is the matching positions, and TP is the target position.

The constant error was calculated as the mean value of the algebraic errors of a test. This means that the constant error represents the average deviation of the matching positions from the target position over all trials.13 The variable error was calculated as the SD of the algebraic errors over the 10 trials, and thus represents the variability of the matching positions.13 The variable error was calculated after detrending the algebraic errors, separately for each of the subject’s test series (ie, least square means was computed for the algebraic errors). This was done to eliminate possible drift in bias, which is unrelated to the response variability, but will affect the variable error.31

The pain drawings were converted to categorical variables by an experienced physiotherapist using the method of Margolis et al.26 The low back areas were defined to the guidelines of the International Association for the Study of Pain,19 which are based on anatomic landmarks (see fig 2). For each of the 6 areas, the pain characteristics were described by 5 dichotomous variables: (1) pain or no pain, (2) area with dominant pain, (3) continuous pain, (4) intermittent pain, and (5) itching pain.

To investigate whether patients who reported high functional disability or low self-efficacy showed larger repositioning errors, patients were divided into subgroups according to whether they scored higher or lower than the median score of ODQ, LBOS, SES, and the number of pain areas, respectively.

We applied the multivariate analysis of covariance (MANCOVA) to assess differences in variable error and constant error between CLBP patients and controls, and between subgroups of patients with CLBP, that is, the subgroups were formed according to medical diagnoses, McKenzie diagnoses, low and high scores on questionnaires (ODQ, SES, LBOS), and according to the number of dominant pain areas. We studied associations between variable error and constant error and the questionnaire scores by calculating 2-tailed partial Pearson correlation coefficients, individually for the 3 measuring points (see fig 1). The distance from the turning angle-to-target position, which is directly related to the ROM, was used as the covariate in both the MANCOVA and correlation analyses.

For reliability, intraclass correlation coefficients (ICCs) were calculated for variable error and constant error in 20 healthy subjects (10 men, 10 women) who executed a slightly modified version of the sitting pelvic test twice, with 15 minutes of rest between tests. The main difference in the pelvic test in the test-retest study was that the back remained in the flexed position a few seconds before the repositioning movement started. We used a 2-way mixed model and absolute agreement for the ICC calculation, and the Kolmogorov-Smirnov test was used to examine the distribution of the data.

The statistical software used was SPSS.c In all analyses, a P value of less than .05 was considered significant.

Back to Article Outline

Results 

The Covariate 

Table 1 shows that there were significant differences between patients and control subjects in ROM and distance between the target position and the turning position (repositioning angle). Because movement distance may influence the variable error, we decided to use the repositioning angle as the covariate in the analyses.

Reliability 

The variable error (.639) and the constant error (.443) were normally distributed (Kolmogorov-Smirnov test), and the ICC with absolute agreement showed a variable error of .754 and a constant error of .860 at S2, indicating acceptable reliability.

Repositioning Acuity 

The distributions of the variable error and the constant error at S2 for 92 patients with CLBP and 31 control subjects are shown in separate box plots in figure 3. There were no differences in variable error and constant error between the patient and the control groups at S2 (variable error, P=.67; constant error, P=.90), T7 (P=.86, P=.57), or at the mid-position (P=1.00, P=.83). There were no differences between men and women in either group.

  • View full-size image.
  • Fig 3. 

    Boxplots showing the distribution of (A) variable errors and (B) constant errors at S2 for patients with CLBP and asymptomatic control subjects. The line inside the boxes indicates the median, the upper and lower limits of the boxes represent the 25th and 75th percentiles, and the whiskers indicate the total range of the data distribution.

Repositioning Acuity and Diagnoses 

There were no differences in the repositioning acuity between subgroups of patients with different diagnoses according to the medical records (S2: variable error, P=.98; constant error, P=.54; T7: variable error, P=.97; constant error, P=.10; mid-position: variable error, P=.86; constant error, P=.36) and to the McKenzie diagnoses (S2: variable error, P=.70; constant error, P=.83; T7: variable error, P=.48; constant error, P=.24; mid-position: variable error, P=.83; constant error, P=.77).

Repositioning Acuity and Self-Estimated Function 

In general, we found only nonsignificant correlations between variable error, constant error, and self-estimated function in the CLBP group. Yet, weak statistically significant positive correlations were observed between variable error and ODQ walking (r=.25, P<.05) and between constant error and ODQ personal care (r=.23, P<.05).

When subgroups of patients with high and low scores on ODQ, SES, LBOS, and number of dominant pain areas were compared with each other and with the control group, we found no differences in repositioning acuity: ODQ (S2: variable error, P=.39; constant error, P=1.00; T7: variable error, P=.94; constant error, P=.79; mid-position: variable error, P=.67; constant error, P=.26); SES (S2: variable error, P=.84; constant error, P=.83; T7: variable error, P=.98; constant error, P=.62; mid-position: variable error, P=.71; constant error, P=.39), LBOS (S2: variable error, P=.74; constant error, P=.86; T7: variable error, P=.97; constant error, P=.64; mid-position: variable error, P=.99; constant error, P=.88); and number of dominant pain areas (S2: variable error, P=.91; constant error, P=.37; T7: variable error, P=.98; constant error, P=.60; mid-position: variable error, P=.57; constant error, P=.60).

Back to Article Outline

Discussion 

In agreement with some studies9, 10, 11 but in disagreement with others,5, 6, 7, 8 our study did not reveal any differences in repositioning errors between patients with CLBP and asymptomatic patients. Neither were there any differences between the control group and subgroups of patients with specific diagnoses, or subgroups with particularly low scores on self-reported function, low self-efficacy, high pain ratings, or widespread pain. Our study was designed to avoid methodologic limitations of previous investigations on lumbar repositioning acuity. First, we included a large sample of patients with well-defined CLBP.5, 6, 7, 9, 10 Second, the repositioning test was standardized with respect to the natural sitting posture and the lumbar ROM, and the experimenter was blinded to the participants’ back pain status.5, 6, 7, 8, 9, 10, 11 Third, cutaneous cues were minimized by the set-up.5, 7, 9, 10, 11 Fourth, confounding effects of differences in movement distance were eliminated.5, 6, 7, 8, 9, 10, 11 Fifth, variable error and constant error were analyzed to allow separation of repositioning variability and bias.5, 6, 8, 9, 10

There may be several reasons for the absence of detectable differences in repositioning errors. One reason might be that the sensitivity of the methods we applied was too poor. It is impossible, for instance, to exclude sensory input from the hip region that to some extent may conceal proprioceptive disturbances of the lumbar region. Yet, there are several reasons to suggest that the methods applied provided a sensitive test of the repositioning acuity of the lumbar spine. Although this possibility is difficult to exclude entirely, it seems unlikely for several reasons. We used a tracking system with high sensitivity to measure the repositioning acuity. The variable error and constant error were measured at 3 positions to permit evaluation of potential differences of repositioning acuity at different levels of the lumbar spine. The calculations of repositioning errors were based on 10 trials intended to improve the stability of the results. Differences in movement distance were controlled for in the statistics. The repositioning test was performed in a sitting posture to avoid potentially biasing effects of afferent input from the legs and feet, and cutaneous cues from equipment, clothes, and hair were minimized.

The test-retest reliability of the error estimates for the pelvic test was found to be acceptable. The ICC results for the variable error and the constant error were similar to those previously reported for absolute error in other lumbar repositioning tests.5, 32 Because it has been shown that the ICC usually is larger for absolute error than the variable error in a given test,11 it is concluded that the sitting pelvic test is at least as reliable as other tests aimed at evaluating the repositioning acuity of the lumbar spine. We therefore consider the power of the study, with such a large sample, to be reasonably good.

In contrast to when only the absolute error is assessed, separate analyses of the variable error and the constant error provide more information on the sensorimotor mechanisms involved in the repositioning task. Thus, the constant error appears to reflect systematic biases in the motor control system while the variable error reflects the accuracy of the position sense.14 This distinction is supported by the observation that the variable error and the constant error were poorly correlated (see also Schmidt and Lee13). For both the error estimates, however, no differences were found between patients and control subjects. The lack of differences between subgroups of rather well-defined patients suggests that lumbar repositioning acuity is poorly related to common diagnostic criteria applied in CLBP (O’Sullivan et al8).

In standing, disturbed lumbar repositioning acuity has been observed on patients with low back pain in some studies5, 11 but not in others (eg, Newcomer et al10). Standing implies, however, that proprioceptive and cutaneous cues from the limbs and feet might have a significant impact on the lumbar repositioning precision. To avoid this, the repositioning test was executed in this study with subjects sitting. The target position of the lumbar spine was defined to deviate from the habitual sitting posture. This was done to avoid a posture that might be reached through automized movements (ie, feed-forward) and, hence, are executed with minimal reliance on proprioceptive input from the lumbar region. It has been indicated that there are larger differences in repositioning acuity between patients with CLBP and asymptomatic subjects after lumbar flexion than after extension or rotations.11 Therefore, the repositioning movement in our study was performed in the sagittal plane after the execution of a lumbar flexion and extension movement.

A methodologic shortcoming of our study is that the repositioning errors were assessed for only 1 target posture with slightly extended lumbar spine. Yet, this approach was chosen to permit a reasonable number of trials for each subject without compromising the subjects’ focus on the task. An alternative would have been to define several target postures and to investigate these with only a few trials each.7, 10, 11 While there is a risk for noisy, and potentially biased, results when calculating an outcome score based on very few trials, the present approach implies that potential effects in other target postures and movement directions passed undetected (eg, Koumantakis et al11).

There is good evidence for sensorimotor disturbances in LBP. Altered coordination patterns of the trunk muscles have been demonstrated in several electromyography studies on patients with LBP (eg, Arendt-Nielsen et al,33 Hodges et al34). It is also well established that mechanoreceptors in muscles, ligaments, disks, and joint capsules can influence the muscle activity through different spinal and supraspinal pathways.35, 36 Thus, altered muscle coordination patterns in CLBP might be caused by disturbed sensory input from the lower back region. Indirect support for this has been presented, for example, by Brumagne et al,37 who found that patients with spinal pain were less disturbed by vibration-induced activation of paraspinal proprioceptors during a postural control test than were asymptomatic subjects.

If patients with CLBP suffer from pain, poor lumbar stability, and muscle coordination,33, 34, 38 but show normal repositioning acuity, then these sensorimotor disturbances must be caused by mechanisms other than through the proprioceptive parameters we analyzed in this study. Biomechanic causes for lumbar pain and instability have been suggested,39 as well as deranged intrinsic motor control mechanisms (eg, feed-forward control of movements).34 Another possibility is that the disturbances are a result of abnormal sensory cues that are not measurable in the repositioning test we used in this study. It is possible that maintenance of normal motor functions of the lumbar spine is dependent on different proprioceptive cues than what are needed for the control of movements of the extremities and the neck. In support for this, animal studies have indicated that the spinal reflex controls of the gamma-muscle spindle system (ie, the main provider of proprioceptive input from the muscles) of lumbar multifidus muscles are quite different from that of muscles of the neck and the extremities.40, 41, 42 It remains to be determined to what extent such differences are related to the fact that the lumbar multifidus muscles, in contrast to muscles of the neck and the extremities, have their primary functions in postural control and as stabilizers of the spine. Regardless of the underlying causes, it seems reasonable to suggest that sensorimotor tests other than lumbar repositioning tests should be used for objective assessment of proprioceptive and motor control deficits in CLBP.

Back to Article Outline

Conclusions 

In this study, we observed only weak (or no) correlations between the error estimates of lumbar repositioning and self-reported functional disability, pain, and self-efficacy. This is perhaps not that surprising because there were no differences between patients and asymptomatic subjects regarding the repositioning acuity. Note also that the self-reported data on functional ability did not contain any direct information on dysfunctions that are suggested to be particularly affected by proprioceptive disturbances such as, for example, clumsiness, poor movement precision, and unsteadiness. Thus, questionnaires aimed at differentiating between different types of sensorimotor deficits would be a valuable tool in future research into potential sensory disturbances related to pain and motor control deficits in CLBP.

The lumbar repositioning acuity in flexion and extension does not appear to differ between patients with CLBP and asymptomatic subjects, nor between groups of patients with various diagnoses and symptoms. We conclude that the proprioceptive ability of the lumbar spine, as assessed in the repositioning test used here, is poorly related to lumbar pain and self-reported dysfunctions. This suggests that sensorimotor functions that are affected in CLBP should be assessed with methods other than the repositioning tests in order to generate objective data on sensorimotor dysfunctions. Such data are needed for the development of rational diagnostic methods and rehabilitation programs.

Suppliers

Back to Article Outline

References 

  1. Revel M , Andre-Deshays C , Minguet M . Cervicocephalic kinesthetic sensibility in patients with cervical pain . Arch Phys Med Rehabil . 1991;72:288–291
  2. Loudon JK , Ruhl M , Field E . Ability to reproduce head position after whiplash injury . Spine . 1997;22:865–868
  3. Kristjansson E , Dall’Alba P , Jull G . A study of five cervicocephalic relocation tests in three different subject groups . Clin Rehabil . 2003;17:768–774
  4. Treleaven J , Jull G , Sterling M . Dizziness and unsteadiness following whiplash injury (characteristic features and relationship with cervical joint position error) . J Rehabil Med . 2003;35:36–43
  5. Gill KP , Callaghan MJ . The measurement of lumbar proprioception in individuals with and without low back pain . Spine . 1998;23:371–377
  6. Newcomer KL , Laskowski ER , Yu B , Johnson JC , An KN . Differences in repositioning error among patients with low back pain compared with control subjects . Spine . 2000;25:2488–2493
  7. Brumagne S , Cordo P , Lysens R , Verschueren S , Swinnen S . The role of paraspinal muscle spindles in lumbosacral position sense in individuals with and without low back pain . Spine . 2000;25:989–994
  8. O’Sullivan PB , Burnett A , Floyd AN , et al.   Lumbar repositioning deficit in a specific low back pain population . Spine . 2003;28:1074–1079
  9. Lam SS , Jull G , Treleaven J . Lumbar spine kinesthesia in patients with low back pain . J Orthop Sports Phys Ther . 1999;29:294–299
  10. Newcomer K , Laskowski ER , Yu B , Larson DR , An KN . Repositioning error in low back pain. Comparing trunk repositioning error in subjects with chronic low back pain and control subjects . Spine . 2000;25:245–250
  11. Koumantakis GA , Winstanley J , Oldham JA . Thoracolumbar proprioception in individuals with and without low back pain (intratester reliability, clinical applicability, and validity) . J Orthop Sports Phys Ther . 2002;32:327–335
  12. Norrsell U , Backlund H , Gothner K . Directional sensibility of hairy skin and postural control . Exp Brain Res . 2001;141:101–109
  13. Schmidt RA , Lee TD . Motor control and learning (a behavioral emphasis) . 3rd ed. Champaign: Human Kinetics; 1999;
  14. Clark FJ , Grigg P , Chapin JW . The contribution of articular receptors to proprioception with the fingers in humans . J Neurophysiol . 1989;61:186–193
  15. Gescheider GA . Psychophysics (the fundamentals) . 3rd ed. Mahwah: Lawrence Erlbaum Associates; 1997;
  16. Descarreaux M , Blouin JS , Teasdale N . Repositioning accuracy and movement parameters in low back pain subjects and healthy control subjects . Eur Spine J . 2005;14:185–191
  17. Riddle DL . Classification and low back pain (a review of the literature and critical analysis of selected systems) . Phys Ther . 1998;78:708–737
  18. McKenzie R . The lumbar spine . Waikanae: Spinal Publications; 2003;
  19. Merskey H , Bogduk N . Classification of chronic pain . Seattle: IASP Pr; 1994;
  20. Riddle DL , Rothstein JM . Intertester reliability of McKenzie’s classifications of the syndrome types present in patients with low back pain . Spine . 1993;18:1333–1344
  21. Kuorinka I , Jonsson B , Kilbom Å . Standardised Nordic questionnaires for the analysis of musculoskeletal symptoms . Appl Ergon . 1987;18:233–237
  22. Borenstein DG . Chronic low back pain . Rheum Dis Clin North Am . 1996;22:439–456
  23. Pearcy MJ , Hindle RJ . New method for the non-invasive three-dimensional measurement of human back movement . Clin Biomech . 1989;4:73–79
  24. Swinkels A , Dolan P . Regional assessment of joint position sense in the spine . Spine . 1998;23:590–597
  25. Hoppenfeldt S . Physical examination of the spine and extremities . Norwalk: Appleton Century Crofts; 1976;
  26. Margolis RB , Chibnall JT , Tait RC . Test-retest reliability of the pain drawing instrument . Pain . 1988;33:49–51
  27. McHorney CA , Ware JE , Lu JF , Sherbourne CD . The MOS 36-item Short-Form Health Survey (SF-36): III. Tests of data quality, scaling assumptions, and reliability across diverse patient groups . Med Care . 1994;32:40–66
  28. Fairbank JC , Couper J , Davies JB , O’Brien JP . The Oswestry low back pain disability questionnaire . Physiotherapy . 1980;66:271–273
  29. Holt AE , Shaw NJ , Shetty A , Greenough CG . The reliability of the Low Back Outcome Score for back pain . Spine . 2002;27:206–210
  30. Denison E , Asenlof P , Lindberg P . Self-efficacy, fear avoidance, and pain intensity as predictors of disability in subacute and chronic musculoskeletal pain patients in primary health care . Pain . 2004;111:245–252
  31. van Beers RJ , Sittig AC , Denier van der Gon JJ . How humans combine simultaneous proprioceptive and visual position information . Exp Brain Res . 1996;111:253–261
  32. Brumagne S , Lysens R , Spaepen A . Lumbosacral repositioning accuracy in standing posture (a combined electrogoniometric and videographic evaluation) . Clin Biomech (Bristol, Avon) . 1999;14:361–363
  33. Arendt-Nielsen L , Graven-Nielsen T , Svarrer H , Svensson P . The influence of low back pain on muscle activity and coordination during gait (a clinical and experimental study) . Pain . 1996;64:231–240
  34. Hodges PW , Moseley GL , Gabrielsson A , Gandevia SC . Experimental muscle pain changes feedforward postural responses of the trunk muscles . Exp Brain Res . 2003;151:262–271
  35. Holm S , Indahl A , Solomonow M . Sensorimotor control of the spine . J Electromyogr Kinesiol . 2002;12:219–234
  36. Sjölander P , Johansson H , Djupsjöbacka M . Spinal and supraspinal effects of activity in ligament afferents . J Electromyogr Kinesiol . 2002;12:167–176
  37. Brumagne S , Cordo P , Verschueren S . Proprioceptive weighting changes in persons with low back pain and elderly persons during upright standing . Neurosci Lett . 2004;366:63–66
  38. Hodges PW , Moseley GL . Pain and motor control of the lumbopelvic region (effect and possible mechanisms) . J Electromyogr Kinesiol . 2003;13:361–370
  39. Panjabi MM . The stabilizing system of the spine. Part II. Neutral zone and instability hypothesis . J Spinal Disord . 1992;5:390–396 ; discussion 397
  40. Johansson H , Djupsjöbacka M , Sjölander P . Influences on the gamma-muscle spindle system from muscle afferents stimulated by KCl and lactic acid . Neurosci Res . 1993;16:49–57
  41. Pedersen J , Sjölander P , Wenngren BI , Johansson H . Increased intramuscular concentration of bradykinin increases the static fusimotor drive to muscle spindles in neck muscles of the cat . Pain . 1997;70:83–91
  42. Kang YM , Wheeler JD , Pickar JG . Stimulation of chemosensitive afferents from multifidus muscle does not sensitize multifidus muscle spindles to vertebral loads in the lumbar spine of the cat . Spine . 2001;26:1528–1536
  • a Polhemus Inc, PO Box 560, Colchester, VT 05446.
  • b Garco R 735; Garco AB, Box 103 265 22 Åstorp, Sweden.
  • c Version 13.0; SPSS Inc, 233 S Wacker Dr, 11th Fl, Chicago, IL 60606.

 Supported by the Svenska Rygginstitutet AB, Sundsvall, Sweden.No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the author(s) or upon any organization with which the author(s) is/are associated.

PII: S0003-9993(06)00472-2

doi:10.1016/j.apmr.2006.05.020

Archives of Physical Medicine and Rehabilitation
Volume 87, Issue 9 , Pages 1170-1176, September 2006

Article Tools

* Image Usage & Resolution