ARTICLE IN PRESS

Manual Therapy 9 (2004) 164172

www.elsevier.com/locate/math

Original article

A survey on the importance of lumbar coupling biomechanics in physiotherapy practice

Chad Cooka,*, Christopher Showalterb
a

Department of Rehabilitation Sciences, Texas Tech University Health Sciences Centre, Odessa, 800 West Fourth Street, Odessa, TX 79763, USA b Maitland Australian Physiotherapy Seminars, P.O. Box 1244, Cutchogue, NY 11935, USA Received 30 June 2003; received in revised form 29 January 2004; accepted 4 March 2004

Abstract Knowledge of lumbar coupling biomechanics is foundational in many manual therapy disciplines. 3-D studies of lumbar coupling indicate that coupling direction may not be predictable. The purpose of this study was to investigate physiotherapists perception of importance of lumbar coupling for validation of manual therapy, necessity in treatment, and perceived direction of lumbar coupling biomechanics. A sample of 369 physiotherapists within the United States volunteered to participate in this study. An ologit regression analysis investigated factors associated with perception of lumbar coupling. A Fleiss Kappa determined the agreement of coupling direction among physiotherapists. The majority of physiotherapists indicated that lumbar coupling biomechanics were important or very important, frequently used during treatment, and necessary for validation of manual therapy. The ologit regressions identied that the importance placed upon lumbar coupling biomechanics is highly related toward therapists preconcept of coupling. Kappa values for the ve spinal segments were negative indicating worse than chance agreement: L122 0:10; L223 0:11; L324 0:09; L425 0:10; and L52S1 0:09: The poor Kappa values, strong pre-conceptual perceptions of coupling necessity, and the importance placed upon lumbar coupling for treatment could lead to disparities among physiotherapists in lumbar manual therapy assessment and treatment. r 2004 Elsevier Ltd. All rights reserved.

1. Introduction Biomechanical analysis, including investigation of coupled motion is often identied as essential in the assessment and treatment of low back pain (Cassidy, 1976; Grice, 1979; Pearcy et al., 1985; Gomez, 1994; Panjabi et al., 1994; Mellin et al., 1995; Winkel et al., 1996). There are two principle components of lumbar coupling; quantity of motion, used in detection of hypo and hypermobility and direction of coupling behaviour. It has been suggested that the link between pathology of the lumbar spine may be best represented by addressing the pattern or direction of coupling behaviour (Fryette, 1954; Faye, 1984; Gertzbein et al., 1986; Plaugher, 1993; Gracovetsky et al., 1995; Winkel et al., 1996; Lund et al., 2002). Others have suggested that the quantity of lumbar coupling behaviour is benecial in assessment and treatment (Pearcy et al., 1985; Panjabi et al., 1994;
*Corresponding author. Tel.: +1-915-335-5370; fax: +1-915-3355365. E-mail address: cookttu@yahoo.com (C. Cook). 1356-689X/$ - see front matter r 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2004.03.003

Harrison et al., 1998). The most controversial of the two assessment methods is the theory of directional lumbar coupling, a theory based on the invalidated premise that a normal lumbar coupling pattern exists in nonpathological individuals (Panjabi et al., 1994; Harrison et al., 1998; Gibbons and Tehan, 2001). Many manual therapy disciplines base specic mobilization, manipulation, and muscle energy techniques on selected theories of lumbar coupling direction (Stoddard, 1972; Beal, 1989; Hartman, 1997). Often, a specic technique requires pre-positioning of for spinal segment apposition or movement into a desired coupling direction. During apposition, the clinician will generally side-bend, adjust the sagittal plane position of the spine, and then rotate to lock the level above or below the targeted segment. These theories are inconsistently reported, lack empirical validation, and are generally dened through expert-based learning models (Harrison et al., 1998; Cook, 2003). The feasibility exists, that if practicing therapists do not share consistent directional coupling pattern expectations, the results of their assessment and treatment may be dissimilar.

ARTICLE IN PRESS
C. Cook, C. Showalter / Manual Therapy 9 (2004) 164172 165

1.1. Coupling denition Coupled motion is the rotation or translation of a vertebral body about or along one axis that is consistently associated with the main rotation or translation about another axis (Panjabi et al., 1992). During movement, translation occurs when movement is such that all particles within that segment move in the same direction with the same velocity (Panjabi et al., 1992). With movement, rotation occurs as a spinning or angular displacement of the vertebral body around some axis. Biomechanical coupling is three-dimensional (3-D) and takes place within six degrees of freedom. The six degrees of freedom can translate along and rotate about each orthogonal axis (Panjabi et al., 1992). The 3-D motions in humans correspond to exion/extension, rotation, and side bending forces; one specic movement initiation (such as side bending) theoretically activates movement in the other ve component motions. The behaviour of the coupled pattern is dependent on the rst motion of initiation (e.g., side bending), the posture of the spine, and the pathology of the segment (Panjabi et al., 1989). 1.2. Coupling measurement and ndings The foundational works on coupling mechanics used observation or controversial two-dimensional (2-D) radiographic imagery (Harrison et al., 1998). Past 2-D studies involved cadaveric tissue, X-rays of live subjects, or single X-rays of segments, and used a small sample of subjects (Evans and Lissner, 1959; Harrison et al., 1998). Prior to 1969, only 2-D studies were executed for spinal coupling, signifying that any study performed prior to 1969 encompassed these errant methods (Harrison et al., 1998). 2-D imagery leads to magnication errors, projection of translations as rotations, rotations as translations, and misleading results (Panjabi et al., 1994; Harrison et al., 1996, 1998). The most frequently used method of 2-D imagery involves lateral stress radiograms, a notoriously unreliable method of analysis that is still commonly used today (Nelson, 1993; Harrison et al., 1999). Basing the coupling pattern on observation and potentially awed 2-D radiographic methods, may be one of the reasons such poor continuity existed among

past researchers. Table 1 outlines early 2-D lumbar coupling theories. Contemporary studies have used three-dimensional (3-D) assessment, which more accurately measures the six degrees of freedom associated with coupling motion (Olin et al., 1976; Rab and Chao, 1977; Stokes et al., 1980). The majority of 3-D coupling studies have investigated coupled lumbar behaviour with side bending initiation while in a neutral postural position (Schultz et al., 1979; Pearcy and Tibrewal, 1984; McGlashen et al., 1987; Panjabi et al., 1989; Vincenzino and Twomey, 1993; Panjabi et al., 1994; Cholewicki et al., 1996). Most studies found that coupling direction was dependent upon the level of the spine more so than whether the segment was from a live or cadaveric specimen (Panjabi et al., 1994). Findings of studies investigating side bending initiated coupled motion in neutral, extension, and exion also display variability and are presented in Table 2. 1.3. Purpose of the study There were two purposes to this study. Firstly, we planned to investigate which variables are associated with the coupling importance in determination of use, validation, and frequency of use during practice. Secondly, to investigate therapists perception of directional coupling biomechanics of the lumbar spine in a non-pathological individual which would identify if clinicians agree on a predetermined coupling direction. The literature is inconsistent in identifying a common lumbar coupling pattern, therefore it was feasible that disparity existed among practicing clinicians. If disparity was found to exist in directional coupling knowledge, and if there was a suggested necessity of lumbar coupling knowledge by many manual therapists, the occurrence of variable assessment and treatment continuity among clinicians was thought to be likely.

2. Materials and methods 2.1. Sample Three hundred and sixty-nine physiotherapists participated in this study. The demographic information

Table 1 Coupled lumbar motion with side bending initiation (neutral spine). Classic observation or 2-D studies Author Lovett (1905) Fryette (1954) Stoddard (1959) Kapandji (1974) Rolander (1966) L12 Same Opposite Opposite Opposite None L23 Same Opposite Opposite Opposite None L34 Same Opposite Opposite Opposite None L45 Same Opposite Opposite Opposite None L5S1 Same Opposite Opposite Opposite None

ARTICLE IN PRESS
166 C. Cook, C. Showalter / Manual Therapy 9 (2004) 164172 Table 2 Coupled lumbar motion with side bending initiation (neutral, extension, and exed spine) 3-D studies on asymptomatic subjects Author Neutral spine Schultz et al. (1979) Pearcy and Tibrewal (1984) McGlashen et al. (1987) Panjabi et al. (1989) Panjabi et al. (1994) Cholewicki et al. (1996) Extended spine Schultz et al. (1979) Vincenzino and Twomey (1993) Panjabi et al. (1989) Flexed spine Vincenzino and Twomey (1993) Panjabi et al. (1989) L12 None None None None or same None Opposite or none L23 None Opposite None Opposite Opposite Opposite or none L34 None Opposite None Opposite Opposite Opposite or none L45 None Opposite None Opposite Opposite Same L5S1 None Opposite or none None Opposite Opposite Opposite

None Opposite Same or none

None Same Opposite

None Opposite Opposite

None Same Opposite

None Opposite Opposite

Same or none Same or opposite

Opposite Opposite

Same Opposite

Opposite Opposite

Same or none Opposite

including background of this sample is presented in Table 3. The physiotherapists were voluntary members of a continuing education course entitled Intermediate Spinal Mobilization, administered at different locations in the contiguous United States. This course was selected because it was the rst course in a manual therapy series and the course participants represented diverse backgrounds. The survey was voluntary and approved by the local University Institutional and Ethics Review Board. 2.2. Procedure The survey was administered during the pre-registration period, prior to the initiation of the coursework. This prevented a pre-biasing regarding lumbar coupling biomechanics and allowed the subject to record their response based on their present knowledge of spinal coupling. The physiotherapists were provided written instructions requesting that they answer the questions to the best of their ability. The survey consisted of 11 questions; the rst four consisting of Likert-type questionnaires associated with (1) How important do you feel the theory of lumbar coupling biomechanics is in your application of manual therapy? (2) How important do you feel lumbar coupling biomechanics is in validating manual therapy? (3) How often do you consider or use lumbar coupling biomechanics during your manual therapy treatment? and (4) How often do you use manual therapy in your treatment of the lumbar spine? Further questions were associated with frequency of manual therapy course attendance, manual therapy background or discipline, educational degree, years of practice experience, age, and gender. Since most manual therapy techniques include side bending initiation methods (Stoddard, 1972; Beal, 1989; Kappler, 1989; Nyberg, 1993; Hartman, 1997), a

Table 3 Demographics of survey participants Age 2029 3039 4049 5059 Gender Male Female Years of experience 05 years 610 years 1120 years Greater than 20 Physical therapy education Bachelors Masters Doctoral Background Cyriax Maitland McKenzie Paris Othernot designated None Osteopathic Kaltenbourne IAOMa Mulligan Grimsby NAIOMTb
a b

144 171 50 4

155 214

218 98 46 7

168 191 10

28 76 126 29 47 33 2 12 3 5 5 3

International Academy of Orthopedic Medicine. North American Institute of Orthopaedic Manual Therapy.

question was selected associated with hypothetical coupling direction during active side bending initiation. The coupling direction question consisted of a blank

ARTICLE IN PRESS
C. Cook, C. Showalter / Manual Therapy 9 (2004) 164172 167

table with a set of ve rows corresponding to the lumbar segments L12 through L5S1, and three sub-rows corresponding to the postural positions of neutral, exion, and extension. Three columns divided all rows outlining the choices of opposite, same, and no coupling. The question please identify which direction you feel the corresponding segment will rotate at each of the designated levels, when active left side bending is initiated was asked, allowing the participants to conceptualise which direction coupling should theoretically occur. This allowed participants to choose the direction of lumbar coupling rotation during active side bending in neutral, exion and extension, for all ve levels of the lumbar spine. Participants were asked to place an X in the column choice marked opposite, same, or no coupling. Since the data are categorical, missing values were not replaced and were recorded as dont know. 2.3. Statistical analysis Because the survey questionnaire was not a preexisting instrument, validation was necessary prior to inference of ndings. Using SPSS version 11.0.1, an exploratory factor analysis with varimax rotation was used to determine if the current format of survey questions were appropriate. The exploratory factor analysis also included the Kaiser Meyer Olin (KMO) statistic, used to determine if the sum of partial correlations are greater than the sum of correlations, the Anti Image Correlation, which measures the strength of the sum of correlations, and Bartletts Test for Sphericity, a test for normalization of the sample and the potential relationship it may yield. Eigenvalues were extracted if greater than 0.5 a cut-off relevant for a sample size of 200 to 300 (Field, 2001). This method would serve to limit the latent variables to those that are most indicative of the variance pool within the study. The use of a factor analysis is appropriate on a sample size of 300 or greater, and is crucial in determining if the latent variables stand up on their own (Field, 2001). Additionally, power for a regression analysis generally required 1015 respondents per question, and required a priori sample size of 220315. Three separate ologit regression analyses were performed (Dependent Variable-1 importance of coupling to their treatment, Dependent Variable-2 importance of coupling to validation of manual therapy, Dependent Variable-3 use of coupling during manual therapy) to determine which independent factors were most inuential to determining coupling biomechanics importance and use. An ologit regression analysis ts ordered logit models of an ordinal (polytomous) dependent variable against the selected independent variables. Independent variables can exist using any form of data; those variables that are nominal and ordinal are dummy

coded allowing individualized association to the dependent variable. Since the dependent variable values consist of ranks without order, the actual values taken on by the dependent variable are irrelevant except that larger values are assumed to correspond to probability of changing one lower Likert choice to one higher outcome, and vice versa (Field, 2001). Since the coupling choices were nominal/categorical, and the sample consisted of a number of raters, a Fleiss Kappa coefcient was used to determine agreement on coupling direction (Uebersax, 1987; Feinstein and Cicchetti, 1990; Portney and Watkins, 2000). According to Fleiss (1981), values exceeding 0.75 suggest strong agreement above chance, values in the range of 0.40 to 0.75 indicate fair levels of agreement above chance, and values 0.40 are indicative of poor agreement above chance levels. SAS version 8.0.1 was used in place of SPSS, since SPSS does not have a multi-rater function for Kappa.

3. Results 3.1. Survey questionnaire validation The survey questionnaire did meet the power and validation requirements for the study. The Kaiser MeyerOlkin was 0.885 (0.5 and above indicates sample adequacy) suggesting that patterns of correlations are relatively compact so component analysis should yield distinct and reliable factors. The Bartletts test for Sphericity was signicant (Pp0:0001; w2 10; 723:1). The Bartletts test indicates that the sample does include relationships between variables, therefore component analysis was appropriate. The survey had no questions that required elimination after assessment of the antiimage correlation; all were above the correlation value of 0.7. The varimax component matrix indicated four main extractions consisting of constructs analogous to Likert-scale questions on coupling importance of use, validation, frequency, and use of manual therapy in the clinic. 3.2. Descriptive ndings Eighty ve point one (85.1%) percent of the participants indicated that lumbar coupling biomechanics was very important or important in their application of manual therapy. Only two participants identied coupling mechanics as denitely not important. Seventy eight point eight (78.8%) percent of participants indicated that the theory of lumbar coupling biomechanics is very important or important in validating manual therapy. Ninety three point two (93.2%) percent of the participants reported they frequently or consistently used lumbar coupling

ARTICLE IN PRESS
168 C. Cook, C. Showalter / Manual Therapy 9 (2004) 164172

biomechanics during their application of manual therapy. Lastly, 88 percent indicated the use of manual therapy frequently, somewhat frequently, or sometimes in the treatment of low back patients. 3.3. Inferential ndings 3.3.1. First ologit regression The model t value for the rst ologit regression was signicant (Pp0:0001; w2 456:2). The Nagelkerke pseudo R2 was 0.71 indicating that the goodness of t within this model is very good. The Nagelkerke pseudo R2 measures the explanatory power of the model, a similar concept to the regression coefcient in a linear model (Long, 1997). The rst ologit regression using the dependent variable how important the theory of lumbar coupling biomechanics is in your application of manual therapy yielded several signicant results. First, the likelihood that one rated the theory of lumbar coupling biomechanics toward the concept of importance was signicantly associated with their frequency of use of coupling biomechanics in a clinical practice (Pp0:0001). The more frequent the consideration the higher the w2 ; indicating a higher assessment of importance. Second, only those who felt that lumbar coupling biomechanics was very important in validation of manual therapy, were likely to consider the approach signicant in their application of manual therapy (P 0:001; w2 10:28). Third, both clinicians with a bachelors (Pp0:0001; w2 13:91) and a masters degree (P 0:002; w2 10:02) considered lumbar coupling biomechanics important in their application of manual therapy. The rst ologit regression results are outlined in Table 4. 3.3.2. Second ologit regression The model t value for the second ologit regression was signicant (Pp0:0001; w2 537:5). The Nagelkerke pseudo R2 was 0.687, also indicating that the goodness of t within this model is also very good. The second ologit regression using the dependent variable how important the theory of lumbar coupling biomechanics is in validating manual therapy yielded many signi-

cant results. First, those who sometimes (P 0:028; w2 4:83) or rarely (P 0:047; w2 3:95) considered lumbar coupling biomechanics in their application of manual therapy did not feel lumbar coupling biomechanics was critical for validation of manual therapy. Second, those who felt lumbar coupling biomechanics was very important (Pp0:0001; w2 21:1) important (P 0:004; w2 8:18) or had no opinion (P 0:027; w2 4:91) of lumbar coupling biomechanics during their application of manual therapy were signicantly likely to report the necessity of validation. Third, those with a bachelors (P 0:008; w2 6:93) and a masters (P 0:024; w2 5:08) level of education did not feel lumbar coupling biomechanics was important in validating manual therapy. Those with a doctorate degree did not achieve signicance. Lastly, those who rarely (P 0:02; w2 4:33) used manual therapy in their application to the spine were more likely to report the necessity of lumbar coupling biomechanics for the validation of manual therapy. The second ologit regression results are presented in Table 5.

3.3.3. Third ologit regression The model t value for the third ologit regression was signicant (Pp0:0001; w2 677:28). The Nagelkerke pseudo R2 was 0.586 indicating that the goodness of t within this model is good. The third ologit regression using the dependent variable how important the theory of lumbar coupling biomechanics is in your application of manual therapy treatment yielded few signicant results. First, those who used manual therapy sometimes, (Pp0:0001; w2 24:72) consistently, 2 (Pp0:0001; w 19:9) and frequently (Pp0:0001; w2 15:8) were likely to report the consideration of lumbar coupling biomechanics during manual therapy treatment. Those who felt lumbar coupling biomechanics were very important, important, somewhat not important and had no opinion in validating manual therapy were all likely to report the consideration or use of coupling during treatment (P 0:001; w2 10:1). The third ologit regression results are presented in Table 6.

Table 4 Ologit Regression 1: How important do you feel the theory of lumbar coupling biomechanics is in your application of manual therapy? Variable Coupling used during manual therapy (Frequently) Coupling used during manual therapy (Consistently) Coupling used during manual therapy (Sometimes) Physical therapy degree (Bachelors) Coupling used during manual therapy (Rarely) Coupling important for validation (Very Important) Physical therapy degree (Masters) B 5.72 5.33 4.37 3.18 3.63 5.04 2.59 SE 1.02 0.987 0.950 0.852 1.06 1.57 0.819 Wald w2 31.46 29.19 21.23 13.94 11.66 10.28 10.02

Indicates level of signicance Pp0.05. Indicates level of signicance Pp0.0001. Only variables that were signicant (Pp0.05) are reported.

ARTICLE IN PRESS
C. Cook, C. Showalter / Manual Therapy 9 (2004) 164172 Table 5 Ologit Regression 2: How important do you feel the theory of lumbar coupling biomechanics is in validating manual therapy? Variable Physical therapy degree (Bachelors) Use of manual therapy in back treatment (Rarely) Physical therapy degree (Masters) Coupling used during manual therapy (Rarely) Coupling used during manual therapy (Sometimes) B 2.28 2.02 1.91 2.01 1.95 SE 0.867 0.885 0.850 1.03 0.888 Wald w2 6.93 5.23 5.08 4.83 3.95 169

Indicates level of signicance Pp0.05. Only variables that were signicant (Pp0.05) are reported.

Table 6 Ologit Regression 3: How often do you consider or use lumbar coupling biomechanics during your manual therapy treatment? Variable Use of manual therapy in back treatment (Frequently) Use of manual therapy in back treatment (Consistently) Use of manual therapy in back treatment (Sometimes) Coupling necessary for validation (Very Important) Coupling necessary for validation (Important) Coupling necessary for validation (No Opinion) Coupling necessary for validation (Somewhat not Important) B 4.14 3.65 3.27 5.14 4.15 3.79 3.57 SE 0.832 0.819 0.818 1.62 1.63 1.65 1.65 Wald w2 24.72 19.90 15.98 10.00 6.46 5.24 4.65

Indicates level of signicance Pp0.05. Indicates level of signicance Pp0.0001. Only variables that were signicant (Pp0.05) are reported.

Table 7 Inter-therapist reliability of perception of coupled motion during side bending initiation Spinal level L12 L23 L34 L45 L5S1 Fleiss Kappa statistic 0.10 0.11 0.09 0.10 0.09

Table 7 outlines the Fleiss Kappa value for each lumbar segment. Spinal segment L12 was 0.10; L23 was 0.11; L34 was 0.09; L45 was 0.10; and L5 S1 was 0.09. A negative Kappa means that there is less agreement than expected by chance given the marginal distributions of ratings.

4. Discussion The ndings of this study indicate that physiotherapists support the importance of lumbar coupling biomechanics for practice use and validation. The majority of the surveyed participants reported that lumbar coupling biomechanics was important in their application of manual therapy. A greater share also reported that lumbar coupling biomechanics is important or very important in validating manual therapy. Nearly all of the participants claimed they used coupling theory at least sometimes during application of manual

therapy. Yet, it is apparent that the directional concept associated with lumbar coupling biomechanics is highly variable among survey participants. The results of each Kappa statistic of each spinal level indicates that there is wide disparity in which coupling theory physiotherapists utilize during biomechanical assessment and treatment of the lumbar spine. Regardless of level, the agreement among therapists for a preconceptual assessment of lumbar coupling motion was poor. A negative Kappa value indicates that the agreement among therapists is actually worse than chance (Landis and Koch, 1977). If the clinicians truly use the directional application of coupling, a concept supported in this surveys ndings (93.2 percent claim use of lumbar coupling) and by the literature (Fryette, 1954; Faye, 1984; Gertzbein et al., 1986; Plaugher, 1993; Gracovetsky et al., 1995; Winkel et al., 1996; Lund et al., 2002), then many clinicians are using conicting conceptual approaches. Since an overwhelming number of survey participants indicate the use of lumbar coupling for treatment, it is likely that selected directional approaches are practiced. Signicance in this study was associated with the preconceptual importance placed upon lumbar coupling biomechanics. Frankly stated, lumbar coupling biomechanics, regardless of which directional model committed to, was used by clinicians who felt it was important for validating and appropriately treating patients. In all three ologit results, with the exception of educational degree, the likelihood that one used, reported consideration of, and felt the importance of

ARTICLE IN PRESS
170 C. Cook, C. Showalter / Manual Therapy 9 (2004) 164172

lumbar coupling biomechanics for validation of manual therapy was highly correlated. This suggests that the direction of lumbar coupling is likely based on a previous assumption, a previous learned model, or a theory that plays a great deal of importance in the application of most manual therapists, despite the fact that physiotherapists cannot seem to agree on which direction that theory best describes lumbar coupling. Strangely, those therapists with both bachelors and masters levels of education reported importance of lumbar coupling biomechanics in their application of manual therapy, though they did not feel lumbar coupling biomechanics was necessary in validating manual therapy. At face value, it appears to be a contrasting concept. There was no trend toward older or younger, or more or less experienced clinicians stating a preference in one coupling pattern or level of importance versus another. Other than degree, it appears any clinician is just as likely to report the importance of lumbar coupling theory regardless of background. The ndings of this study somewhat replicate the ndings of 3-D lumbar coupling biomechanical literature, which for several years has presented disparate directional ndings in lumbar coupling research (Schultz et al., 1979; Pearcy and Tibrewal, 1984; McGlashen et al., 1987; Panjabi et al., 1989; Vincenzino and Twomey, 1993; Panjabi et al., 1994; Cholewicki et al., 1996). Recent 3-D analyses of coupling motion support the theory that lumbar coupling direction is unpredictable, even during analysis of nonpathological subjects. The affect of an injury further complicates the ndings of coupled motion with recent studies reporting variances, in both quantity and directional coupling behaviour (Panjabi et al., 1984; Gertzbein et al., 1985, 1986; Parnianpour et al., 1988; Gracovetsky et al., 1995; Kaigle et al., 1995, 1998; Lund et al., 2002). No studies were found that identify a predictable, pathological lumbar coupling pattern that correlates with a specic pathology. The ability to predict pathological coupling patterns during assessment would be dependent on repeatable coupling consistency between pathological and non-pathological subjects. To designate a motion as abnormal spinal coupling requires knowledge of the normal motions (Panjabi et al., 1994). Parameters of normal spine coupling are unavailable (Harrison et al., 1998). The possibility exists that educational programs and continuing education groups are teaching selective, conicting results from updated 3-D studies on directional-coupled motion (Cook et al., 2002). This could account for the inconsistency of ndings, specically associated with the poor Kappa values. Despite this possibility, there were no backgrounds (i.e., Maitland, McKenzie, etc.) that were signicantly associated with lumbar coupling. Though not signicant (a 0:05) all backgrounds reported a positive beta estimate, indicat-

ing an assumed importance of lumbar coupling biomechanics in the treatment, validation, and assessment of low back patients. The data did not dene one background that placed a greater amount of importance in lumbar coupling versus another. This studys ndings support culminating evidence that basing manual therapy techniques on the assumption that a single directional coupling pattern exists for all patients is erroneous. Pertinent lumbar spine coupling assessment appears to depend on the anatomical differences, posture and pathology of each individual. Since many coupling movements are on the order of 11.1 mm per segment this form of assessment may not be clinically signicant (Panjabi et al., 1994). Directional coupling assessment may be beyond the skilled practitioners ability to feel the nite movements (Harrison et al., 1998) and since pathologies do not yield specic directional ndings, the information obtained may actually yield little clinical value (Panjabi et al., 1994). Coupling motion assessment may not provide the manual therapist with accurate or useful information unless the pattern of coupling ts that particular patient (Harrison et al., 1998). 4.1. Limitations of the study The ndings of this survey were drawn from participants of one manual therapy continuing education programme. This approach does not rely heavily on directional coupling patterns of the lumbar spine for assessment and treatment. The possibility exists, though unlikely according to the level of importance most clinicians placed on coupling biomechanics, that the participants of the course selected the manual therapy continuing education course based on this reason. Nearly 95 percent of the survey respondents indicated coupling motions were important in their application of manual therapy, a gure that does not support a nonadherence to coupling principles. Additionally, it is likely that many of the course participants categorized their approach on a background or discipline in which they were not versed or skilled in. Many indicated that their approach did not follow one specic background and selected other. The possibility exists, that ones exposure to a particular manual therapy model included incorrect or misinformation, thus reducing the purity of this model. Lastly, these ndings are limited to United States clinicians; generalization to non-US clinicians is unwise.

5. Conclusion This study found that a sample of practicing clinicians participating in a manual therapy course place a high level of importance on lumbar coupling biomechanics,

ARTICLE IN PRESS
C. Cook, C. Showalter / Manual Therapy 9 (2004) 164172 171

but did not agree on the direction of coupled motion of lumbar spine. The poor agreement is not indicative of background differences, age, years of experience, gender, or exposure to past manual therapy courses. The importance of lumbar coupling biomechanics for use, validation and frequency seems to be predicated on the internal importance of lumbar coupling conceptualised by each clinician. Indeed, lumbar coupling associated with quantity of motion for detection of hypo and hypermobility may be benecial. Yet, the majority of lumbar coupling assessment methods involve directional analysis of coupling patterns. In this study, each clinicians concept of lumbar coupling direction is not consistent. Future research should investigate whether clinicians can accurately detect coupled motion as determined by 3-D research. Additionally, future investigation whether coupling ndings in the clinic are reective of preconceived coupling motion is worth exploration.

References
Beal M. Teaching of basic principles of osteopathic manipulative techniques. In: Beal MC, editor. The principles of palpatory diagnosis and manipulative technique. Newark, NJ: American Academy of Osteopathy; 1989. p. 1164. Cassidy J. Roentgenological examination of the functional mechanics of the lumbar spine in lateral exion. Journal of the Canadian Chiropractic Association 1976;20:136. Cholewicki J, Crisco J, Oxland T, Yamamoto I, Panjabi M. Effects of posture and structure on three-dimensional coupled rotations in the lumbar spine. Spine 1996;21(21):24218. Cook C. Lumbar coupling biomechanicsa literature review. Journal of Manual and Manipulative Therapy 2003;11(3):13745. Cook C, Turney L, Ramirez L, Miles A, Haas S, Karakostas T. Predictive factors in poor inter-rater reliability among physical therapists. Journal of Manual and Manipulative Therapy 2002;10(4):2005. Evans F, Lissner H. Biomechanical studies on the lumbar spine and pelvis. American Journal of Bone and Joint Surgery 1959;41: 27890. Faye LJ. Motion palpation of the spine. Huntington Beach: Motion Palpation Institute; 1984. Feinstein A, Cicchetti D. High agreement but low kappa: I. the problems of two paradoxes. Journal of Clinical Epidemiology 1990;43(6):5439. Field A. Discovering statistics using spss for windows. London: Sage Publications, 2001. p. 42552. Fleiss JL. Statistical methods for rates, proportions, 2nd ed. New York: John Wiley & Sons; 1981. p. 28491. Fryette H. The principles of osteopathic technique. Carmel CA: Academy of Applied Osteopathy; 1954. p. 114. Gertzbein S, Seligman J, Holtby R. Centrode patterns and segmental instability in degenerative disc disease. Spine 1985;14:594601. Gertzbein S, Seligman J, Holtby R. Centrode characteristics of the lumbar spine as a function of segmental instability. Clinical Orthopedics 1986;208:4851. Gibbons P, Tehan P. Spinal manipulation: indications, risks and benets. Journal of Bodywork and Movement Therapies 2001; 5(2):1109.

Gomez T. Symmetry of lumbar rotation and lateral exion range of motion and isometric strength in subjects with and without back pain. Journal of Orthopedic and Sports Physical Therapy 1994;19:428. Gracovetsky S, Newman N, Pawlowsky M, Lanzo V, Davey B, Robinson L. A database for estimating normal spinal motion derived from noninvasive measurements. Spine 1995;20(9): 103646. Grice A. Radiographic, biomechanical and clinical factors in lumbar lateral exion: Part 1. Journal of Manipulative Physiological Therapeutics 1979;2:2634. Harrison D, Harrison D, Troyanovich S. Three-dimensional spinal coupling mechanics: Part one. Journal Manipulative Physiological Therapeutics 1998;21(2):10113. Harrison D, Harrison D, Troyanovich S. In Plaughers; Threedimensional spinal coupling mechanics: Part 1A review of the literature. Journal of Manipulative Physiological Therapeutics 1999;22:5. Harrison D, Harrison D, Troyanovich S, Hansen D. The anterior posterior full-spine view: the worst radiographic view for determination of mechanics of the spine. Chiropractic Technique 1996;8:16370. Hartman L. Handbook of osteopathic technique. 3rd ed. San Diego, CA: Singular Pub Group; 1997. p. 1313. Kaigle A, Holm S, Hansson T. Experimental instability in the lumbar spine. Spine 1995;20:42130. Kaigle A, Wessberg P, Hansson T. Muscular and kinematic behavior of the lumbar spine during exion-extension. Journal of Spinal Disorders 1998;11:16374. Kapandji I. Physiology of the joints: the trunk and the vertebral column, 2nd ed. Vol. 3. London: Churchill Livingstone; 1974. p. 11447. Kappler R. Direct action techniques. In: Beal MC, editor. The principles of palpatory diagnosis and manipulative techniqu. Newark: American Academy of Osteopathy; 1989. p. 1164. Landis J, Koch G. The measurement of observer agreement for categorical data. Biometrics 1977;33:15974. Long JS.Regression models for categorical, limited dependent variables, 7th ed. London: Sage Publications; 1997. p. 260. [Chapter 8]. Lund T, Nydegger T, Schlenzka D, Oxland T. Three-dimensional motion patterns during active bending in patients with chronic low back pain. Spine 2002;27(17):186574. McGlashen K, Miller J, Schultz A, Andersson G. Load displacement behavior of the human lumbo-sacral joint. Journal of Orthopedic Research 1987;5:48898. Mellin G, Harkapaa K, Hurri H. Asymmetry of lumbar lateral exion and treatment outcome in chronic low back pain patients. Journal of Spinal Disorders 1995;8:159. Nelson C. The cognitive roots of chiropractic theories and techniques. Journal of Chiropractic Humanities. 1993;1:4255. Nyberg R. Manipulation: denition, types, application. In: Basmajian J, Nyberg R, editors. Rational Manual Therapie. Baltimore, MD: Williams and Williams; 1993. p. 2148. Olin T, Olsson T, Selvik G, Willner S. Kinematic analysis of experimentally provoked scoliosis in pigs with Roentgen stereophotogrammetry. Acta Radiological Diagnosis 1976;1(1):10727. Panjabi M, Hult J, Crisco J, White A. Biomechanical studies in cadaveric spines. In: Jayson M, editor. The Lumbar Spine and Back Pain. 4th ed. London: Churchill Livingstone; 1992. p. 13355. Panjabi M, Krag M, Chung T. Effects of disc injury on mechanical behavior of the human spine. Spine 1984;9(7):70713. Panjabi M, Oxland T, Yamamoto I, Crisco J. Mechanical behavior of the human lumbar and lumbosacral spine as shown by threedimensional loaddisplacement curves. American Journal of Bone and Joint Surgery 1994;76:41324.

ARTICLE IN PRESS
172 C. Cook, C. Showalter / Manual Therapy 9 (2004) 164172 Rolander S. Motion of the lumbar spine with special reference to the stabilizing effect of posterior fusion. Acta Orthopedic Scandanavia Supplement 1966;90:1144. Schultz A, Warwick D, Berkson M, Nachemson A. Mechanical property of the human lumbar spine motion segments; part 1. Responses in exion, extension, lateral bending and torsion. Journal of Biomedical Engineering 1979;101:4652. Stokes I, Medlicott P, Wilder D. Measurement of movement in painful intervertebral joints. Medical Biological Engineering & Computers 1980;18:694700. Stoddard A. Manual of osteopathic technique. 2nd ed. London: Hutchinson Medical Publications; 1972. p. 268. Uebersax J. Diversity of decision making models and the measurement of interrater agreement. Psychological Bulletin 1987;101:1406. Vincenzino G, Twomey L. Side-exion induced lumbar spine conjunct rotation and its inuencing factors. Australian Journal of Physiotherapy 1993;39(4):299305. Winkel D, Aufdemkampe G, Matthijs O, Phelps V. Diagnosis and treatment of the spine. Gaithersburg, MD: Aspen Publication; 1996. p. 1808. Panjabi M, Yamamoto I, Oxland T, Crisco J. How does posture affect coupling in the lumbar spine. Spine 1989;14(9):100211. Parnianpour M, Nordin M, Frankel V, Kahanovitz N. Trunk triaxial coupling of torque generation of trunk muscles during isometric exertions and the effect of fatiguing isoinertial movements on the motor output and movement patterns. Spine 1988;13: 98292. Pearcy M, Portek I, Shepherd J. The effect of low back pain on lumbar spine movements measured by three-dimensional X-ray analysis. Spine 1985;10:1503. Pearcy M, Tibrewal M. Axial rotation and lateral bending in the normal lumbar spine measured by three-dimensional radiography. Spine 1984;9(6):5827. Plaugher G. Textbook of clinical chiropractic: A specic biomechanical approach. Baltimore, MD: Williams & Wilkins; 1993. p. 1525. Portney L, Watkins M. Foundations of clinical research. 2nd ed. Upper Saddle River, New Jersey: Prentice-Hall; 2000. p. 56165. Rab G, Chao E. Verication of roentgenographic landmarks in the lumbar spine. Spine 1977;2:28793.