Free
Original Contribution  |   July 2012
Cervical Spine Bending: A Factor Confounding Whole Trunk and Lumbar Forward Bending Range of Motion
Author Notes
  • From the Restorative Care Foundation in Kansas City, Missouri (Drs Brooks and Patterson), the Kansas City University of Medicine and Biology in Missouri (Drs Brooks and Wagner), and the Nova Southeastern University College of Osteopathic Medicine (NSU-COM) in Ft Lauderdale, Florida (Drs Patterson [retired] and Hardigan). All research was conducted at NSU-COM 
  • Address correspondence to William J. Brooks, DO, Restorative Care Foundation, 9204 NW 80th Terr, Kansas City, MO 64152-1616. E-mail: wjbdo@wjbrooksdo.com  
  •    Editor's Note: A supplemental video that depicts the study procedure for the present article is available online at http://www.jaoa.org/content/112/7/429/suppl/DC1.
     
Article Information
Emergency Medicine / Pulmonary Disorders
Original Contribution   |   July 2012
Cervical Spine Bending: A Factor Confounding Whole Trunk and Lumbar Forward Bending Range of Motion
The Journal of the American Osteopathic Association, July 2012, Vol. 112, 429-436. doi:10.7556/jaoa.2012.112.7.429
The Journal of the American Osteopathic Association, July 2012, Vol. 112, 429-436. doi:10.7556/jaoa.2012.112.7.429
Abstract

Context: Knee bending during tests of lumbar forward bending (FB) may introduce confounding variability. Precluding bending at the knees has, therefore, long been standard protocol to produce valid and reproducible results. However, there is limited research on cervical spine bending as a confounding variable in whole trunk and lumbar FB.

Objective: To examine the role of cervical spine bending on the range of whole trunk and lumbar FB.

Methods: Participants were recruited from the faculty, staff, and student population of Nova Southeastern University's Health Professions Division. Each participant underwent 4 FB tests with varying cervical starting positions. Range of motion was measured for whole trunk FB and lumbar FB by using the fingertip-to-floor and double digital inclinometer techniques, respectively.

Results: Two hundred thirty-six participants met the study criteria. Statistically significant differences were found in both whole trunk (6.96 cm) and lumbar (3.95°) FB range of motion when the cervical spine was backward bent after full spine FB (P<.05). Statistically significant differences were also found in both whole trunk (15.72 cm) and lumbar (7.38°) FB when the cervical spine was backward bent before thoracolumbar spine FB (P<.05).

Conclusion: Cervical spine bending influences the ability of the trunk and lumbar spine to bend forward and is, therefore, a confounding variable during tests of whole trunk and lumbar spine FB.

Low back pain is well recognized as an enormous cost to society both in direct health care expense and in being the most frequent cause of disability in working-age adults. Manipulative medicine has been recognized as an effective treatment for patients with low back pain.1,2 Osteopathic manipulative medicine is understood, in part, as a means to restore available range of motion at a joint. Thus, at least 1 benefit of manipulative medicine for low back pain has been hypothesized to be the restoration of spinal mobility.3,4 However, published studies5-20 continue to reveal confusing evidence about what relationships exist between lumbar motion and low back pain syndromes. 
An array of techniques for measuring spinal motion continues to be explored with reference to validity, reliability, safety, and practicality.17,21-31 An unrecognized confounding variable would undermine these efforts. Indeed, the results of these efforts have been sufficiently problematic—especially with regard to interexaminer reliability—such that spinal range of motion is no longer recognized as a criterion for impairment ratings by the American Medical Association (AMA).32 In spite of these limitations, lumbar range of motion continues to be used as a fundamental indicator of function for clinical evaluation.33-36 
In the present study, we focused specifically on forward bending (FB) of the spine in the standing context. Interexaminer reliability has been difficult to establish for standing lumbar FB.22,37-41 Research on potential confounding variables has focused on the influence of age, sex, time of day, warm-up or no warm-up, motion at the hips, and, recently, motion at the knees and ankles.12,42-45 
Until the 1980s, lumbar FB range of motion was commonly measured with the fingertip-to-floor test. However, the fingertip-to-floor test came to be appreciated as an invalid (construct) measure for between-subjects comparison of lumbar motion because anatomic variations of the cephalad extremities and thoracic region as well as motion throughout the thoracic region and at the hips are confounding variables. The fingertip-to-floor test is, therefore, also an invalid test of lumbar FB range of motion for within-subject comparison because of motion at the hips and of the thoracic spine. The current AMA-sanctioned standard for measuring spinal region motion is the dual inclinometer technique.46 For measurement of lumbar motion, the method subtracts motion at the hips, allowing for appreciation of lumbar motion specifically. The sacral inclinometer reading—which is presumed to represent FB at the hips—is subtracted from the inclinometer reading at L1. Although dual inclinometer technique isolates measurement between specific loci, full FB of the lumbar spine includes FB of the thoracic spine and at the hips. To appreciate this total motion, the fingertip-to-floor test remains a construct valid method for within-subject comparisons of whole-trunk FB. 
There are conflicting indirect data on the influence at the knee and ankle bending during tests of lumbar FB range of motion.6,23,45,47-50 Standard AMA protocol requires the standing patient to maintain extension at the knees to control for the possibility that flexion at the knees introduces confounding variability.46 Our literature search did not uncover a single exception to this requirement in all experimental protocols of standing FB. Notably, however, the literature search also did not uncover any study that specifically investigated the confounding influence of knee bending. 
The possible influence of knee bending on FB range of motion led the primary investigator (W.J.B.) to question whether neck bending influences range of motion in whole trunk FB and lumbar FB. Clinical observations did suggest an influence, which is consistent with the osteopathic tenet that the body is a unit.51 We found 2 reports52,53 that focused on the contribution of structures cephalad to the lumbar spine during FB. One reported that in baboon cadavers there was slight displacement of the spinal cord when the cervical spine and/or the thigh with extension at the knee was forward bent.52 The study did not address lumbar or trunk range of motion. Another study concluded that the human caudal thoracic spine forms a functional unit with the lumbar spine during whole trunk FB.53 We found only 1 experimental protocol in which each participant was asked to maintain a specific position of the neck (forward bent) during measurement of lumbar FB range of motion. Notably, cervical FB was not monitored in that study.29 
The goal of the present study was to examine the role of cervical spine bending on the range of whole trunk and lumbar FB. Two null hypotheses were formulated: (1) with extension at the knees, cervical spine bending would have no effect on lumbar FB range of motion, and (2) with extension at the knees, cervical spine bending would have no effect on whole trunk FB range of motion. 
To test these null hypotheses, 3 tests of FB were designed, including unspecified FB (test A), specified sequential cephalic to caudal FB (test B1), sustained whole trunk FB with specified cervical backward bending (BB) followed by further whole trunk FB (test B2), and specified sequential cephalic to caudal FB with cervical BB throughout the movement (test C). 
We predicted that test A would result in greater FB range of motion than test B1. We also predicted that test B2 would allow for additional range of motion in lumbar FB and whole trunk FB. Finally, we predicted that test A and test B2 would produce similar findings. As we had no clinical experience with test C, we had no expectations regarding test C. 
Methods
Participants
Participants were recruited from personnel of the Nova Southeastern University Health Professions Division in February and March 2002. Recruitment took place on each day of testing by asking passing students, faculty, and staff to participate. There were a total of 3 days of testing (2 in February and 1 in March). Participants gave written informed consent to participate in the study, which was approved by the Nova Southeastern University Institutional Review Board. Inclusion criteria were ages 18 to 65 years, ability to stand and bend forward, no current severe illness, and no recent surgery. Patients were excluded if they could touch past the floor while bending forward. 
Method
Three tests (A, B1, B2, and C) were developed to measure FB. A video of these tests is available online at http://www.jaoa.org/content/112/7/429/suppl/DC1. Before formal data collection was performed, all examiners underwent training the day prior to testing, during which they practiced on participants who gave informed consent and were blind to the hypotheses. Training data were collected, and subsequent debriefing and further practice were supervised by W.J.B. 
All tests took place mid-morning through afternoon. Participants were blind to the hypotheses and expectations of the study. Sex, height, age, current neck or back pain, history of spinal surgery, and medical intervention for neck or back pain were recorded for each participant. Before performing the tests, each participant removed his or her shoes, stood erect on a 4-inch raised platform, and looked straight ahead. A dual digital inclinometer (Saunders Digital Inclinometer; The Saunders Group, Inc; Chaska, Minnesota) was placed before testing. Examiner 3 found the S2 vertebra by palpating medially from the posterior superior iliac spines, marked that point with black marker, and placed 1 digital inclinometer below the point. Examiner 4 located the iliac crests, palpated medially to the L4 spinous process and then cephalad to the L1 spinous process, and marked and then placed an inclinometer over L1. Each inclinometer was held at the same location throughout all tests and was zeroed before each test with the participant standing erect. 
Test A was performed first. Test B1 or C was performed after test A, with the order alternated randomly between participants. Test B2 was performed immediately after test B1. Unlike test A, tests B1, B2, and C were guided with the direction and sequencing of neck bending specified and carefully maintained by tactile monitoring. In all tests, extension was maintained at the knees. Examiner 5 (W.J.B.) provided verbal and tactile instruction from each participant's right side. Examiner 1 monitored the participant's knees and any effort to bend at the knees during all 3 tests. For each test, examiner 2 measured the fingertip-to-floor distance with a modified skirt ruler to assess whole trunk FB. Dual digital inclinometer method was used to measure lumbar region FB. The inclinometer at L1 measured FB at L1. The sacral inclinometer measured FB at the hip. Lumbar region FB was calculated by subtracting FB at the hips from FB at L1. 
Test A—The first test always consisted of nonspecified FB, similar to current AMA protocol. Examiner 5 instructed each participant to “please bend forward as far as you can in whatever way is natural, letting your arms dangle in front of you and keeping your fingers straight.” 
Test B1For test B1, or progressive segmental FB, examiner 5 instructed each participant as follows: “Following my guidance, please bend sequentially forward, beginning with your head.” Examiner 5 placed his right hand on the participant's occiput and monitored the head position throughout the test. With his left hand, he provided tactile cues to the participant by running his fingertips from the head sequentially caudad along the spinous processes to the sacrum. As cervical FB progressed, he added, “Keep your chin tucked. Let your arms dangle in front of you, keeping your fingers straight.” If the participant did not maintain FB of the head and neck, the instructions and the test were repeated once. 
Test B2Test B2, or progressive segmental FB followed by specified cervical backward bending (BB), followed by further trunk FB, invariably began from the position of the participant at the conclusion of test B1. Examiner 5's left hand was placed on the participant's right cephalic thoracic region and his right hand on the participant's forehead. Examiner 5 instructed the participant to “please remain forward bent, except allow your head and neck to bend backward.” Examiner 5 tactilely guided the participant's neck into complete and sustained BB and monitored the head position throughout the test. The cephalic thoracic region was monitored by the instructor's left hand to ensure that there was no BB of the thoracic spine. He then asked the participant to “please bend further forward as far as you are able.” 
Test C—For test C, or “specified cervical BB followed by whole trunk FB,” examiner 5's left hand was placed on the participant's right cephalic thoracic region and the fingertips of his right hand were placed on the forehead of the participant, who was asked to “please bend just your neck backward.” Examiner 5 guided the participant's neck into complete BB and monitored the forehead throughout the remainder of the test. Backward bending of the thoracic spine was monitored with his left hand. If BB progressed into the thoracic or lumbar spine regions, the participant was returned to the starting position and the process was repeated once. The participant was then instructed, “Now please bend the rest of your spine forward as far as you can, beginning here (T1) and progressing down your spine. Let your arms dangle in front of you, keeping your fingers straight.“ Simultaneously, examiner 5 ran the fingertips of his left hand sequentially caudad along the spinous processes from T1 to the sacrum. 
Each test was considered complete when FB ceased or when the participant could no longer maintain the specified position of the neck or knees. 
Data Recording
The fingertip-to-floor distance and inclinometer readings for each test were recorded on a numbered data sheet that also contained each participant's sex, height, weight, age, current neck or back pain, history of spinal surgery, and medical intervention for neck or back pain. 
Statistical Analysis
Generalized estimating equations were used to assess the differences in FB while controlling for the covariates gender, age, height, weight, and current pain (neck or back). The generalized estimating equations model used the Gaussian distribution with an independent correlation structure. Generalized estimating equations are methods of parameter estimation for correlated data. If these correlations are not taken into account, the standard errors of the parameter estimates will be invalid and hypothesis-testing results nonreplicable. Descriptive statistics were calculated for the covariates. An α level of .05 was set for all statistical significance testing. 
Results
Two hundred ninety-six participants were recruited (219 in February 2002; 77 in March 2002). Sixty participants who did not report for data collection, had undergone spinal surgery, bent past touching the floor, or had data input errors were excluded. Two hundred thirty-six participants were consequently included in the data analysis. A complete data set, which was used for the generalized estimating equations model, was available for 232 participants. Participants' mean (standard deviation [SD]) age was 33.4 (12.6) years. One hundred seventy-five (75%) participants were not experiencing pain at the time of the experiment (Table 1). Descriptive statistics are presented in Table 1. To increase study power, we analyzed participant demographic variables between the February and March trials (data from the 2 February trials were combined for this analysis). Results of the χ2 analysis and the independent t tests indicated no statistically significant differences between the 2 groups. Therefore, we pooled the data for the generalized estimating equations. 
Table 1.
Descriptive Statistics of Participants in a Study on Cervical Spine Bending and Forward Bending
Variable Mean (SD) Range
Age, ya 33.4 (12.6) 21.0-75.0
Height, inb 67.2 (4.4) 59.0-83.0
Weight, lbc 162.6(35.5) 90.0-260.0
Sexd
Male 123 (52.6) NA
Female 111 (47.4) NA
Paind
Y 59 (25.2) NA
N 175 (74.8) NA
  a n=236
  b n=235
  c n=236
  d Data presented as No. (%)
  Abbreviations: NA, not applicable; SD, standard deviation
Table 1.
Descriptive Statistics of Participants in a Study on Cervical Spine Bending and Forward Bending
Variable Mean (SD) Range
Age, ya 33.4 (12.6) 21.0-75.0
Height, inb 67.2 (4.4) 59.0-83.0
Weight, lbc 162.6(35.5) 90.0-260.0
Sexd
Male 123 (52.6) NA
Female 111 (47.4) NA
Paind
Y 59 (25.2) NA
N 175 (74.8) NA
  a n=236
  b n=235
  c n=236
  d Data presented as No. (%)
  Abbreviations: NA, not applicable; SD, standard deviation
×
Fingertip to Floor (Whole Trunk FB)
Controlling for sex, age, height, weight, and current pain (neck or back), statistically significant differences (P<.05) were found between tests C and B2, between tests C and A, between tests C and B1, between tests B1 and B2, and between tests B1 and A. Test C resulted in the least FB range of motion, with a mean (SD) fingertip-to-floor distance of 50.18 (2.58) cm (95% confidence interval [CI], 48.87-51.48). Test B2 resulted in the greatest FB range of motion with a mean (SD) fingertip-to-floor distance of 34.43 (2.65) cm (95% CI, 33.13-35.74) (Table 2 and Table 3). 
Table 2.
Forward Bending Range of Motion With Cervical Spine Bending (N=232)
Test Mean (SD) FB Range of Motion 95% CI
Whole-Trunk FB, cm
A 36.63 (2.57) 35.32-37.94
B1 41.45 (2.66) 40.14-42.76
B2 34.43 (2.65) 33.13-35.74
C 50.18 (2.58) 48.87-51.48
Sacrum FB,°
A 46.9 (0.7) 45.3-48.6
B1 41.3 (0.6) 39.7-42.9
B2 47.5 (0.6) 45.9-49.2
C 41.3 (0.6) 39.7-43.0
L1 FB,°
A 94.4 (4.0) 92.1-96.8
B1 87.3 (4.2) 84.9-89.7
B2 97.5 (4.4) 95.1-99.9
C 83.9 (4.4) 81.5-86.3
Lumbar FB,°
A 47.5 (0.9) 45.6-49.4
B1 46.0 (0.8) 44.1-47.9
B2 50.0 (0.8) 48.0-51.9
C 42.6 (0.8) 40.7-44.5
  Abbreviations: CI, confidence interval; FB, forward bending; SD, standard deviation.
Table 2.
Forward Bending Range of Motion With Cervical Spine Bending (N=232)
Test Mean (SD) FB Range of Motion 95% CI
Whole-Trunk FB, cm
A 36.63 (2.57) 35.32-37.94
B1 41.45 (2.66) 40.14-42.76
B2 34.43 (2.65) 33.13-35.74
C 50.18 (2.58) 48.87-51.48
Sacrum FB,°
A 46.9 (0.7) 45.3-48.6
B1 41.3 (0.6) 39.7-42.9
B2 47.5 (0.6) 45.9-49.2
C 41.3 (0.6) 39.7-43.0
L1 FB,°
A 94.4 (4.0) 92.1-96.8
B1 87.3 (4.2) 84.9-89.7
B2 97.5 (4.4) 95.1-99.9
C 83.9 (4.4) 81.5-86.3
Lumbar FB,°
A 47.5 (0.9) 45.6-49.4
B1 46.0 (0.8) 44.1-47.9
B2 50.0 (0.8) 48.0-51.9
C 42.6 (0.8) 40.7-44.5
  Abbreviations: CI, confidence interval; FB, forward bending; SD, standard deviation.
×
Table 3.
Cervical Spine Bending and Forward Bending (FB): Generalized Estimating Equations Model With Covariates (N=232)
Test Mean Difference Between Tests (95% Confidence Interval)
Lumber FB Test B1 B2 C
A 1.5 (-2.1-5.1) 2.5 (-1.1-6.0) 4.93 (1.4-8.5)a
B1 4.0 (0.4-7.5)a 3.4 (-0.1-7.0)
B2 7.4 (3.8-11)a
Sacrum FB Test
A 5.7 (2.8-8.6)a 0.5 (-2.4-3.4) 5.5 (2.6-8.4)a
B1 6.2 (3.3-9.1)a 0.1 (-2.8-3.0)
B2 6.1 (3.2-9.0)a
L1 FB
A 7.1 (2.7-11.6)a 3.1 (-1.4-7.5) 10.5 (6.1-15.00)a
B1 -10.2 (5.8-14.6)a 3.4 (-1.0-7.8)
B2 13.6 (9.2-18.0)a
Whole-Trunk FB
A 4.82 (2.39-7.26)a 2.13 (-0.30-4.57) 13.59 (11.15-16.03)a
B1 6.96 (4.52-9.40)a 8.76 (6.33-11.20)a
B2 15.72 (13.29-18.16)a
  a Difference was statistically significant (P<.05).
Table 3.
Cervical Spine Bending and Forward Bending (FB): Generalized Estimating Equations Model With Covariates (N=232)
Test Mean Difference Between Tests (95% Confidence Interval)
Lumber FB Test B1 B2 C
A 1.5 (-2.1-5.1) 2.5 (-1.1-6.0) 4.93 (1.4-8.5)a
B1 4.0 (0.4-7.5)a 3.4 (-0.1-7.0)
B2 7.4 (3.8-11)a
Sacrum FB Test
A 5.7 (2.8-8.6)a 0.5 (-2.4-3.4) 5.5 (2.6-8.4)a
B1 6.2 (3.3-9.1)a 0.1 (-2.8-3.0)
B2 6.1 (3.2-9.0)a
L1 FB
A 7.1 (2.7-11.6)a 3.1 (-1.4-7.5) 10.5 (6.1-15.00)a
B1 -10.2 (5.8-14.6)a 3.4 (-1.0-7.8)
B2 13.6 (9.2-18.0)a
Whole-Trunk FB
A 4.82 (2.39-7.26)a 2.13 (-0.30-4.57) 13.59 (11.15-16.03)a
B1 6.96 (4.52-9.40)a 8.76 (6.33-11.20)a
B2 15.72 (13.29-18.16)a
  a Difference was statistically significant (P<.05).
×
Sacrum Inclinometer (FB of the Trunk at the Hip)
Controlling for sex, age, height, weight, and current pain (neck/back), statistically significant differences (P<.05) were found between tests B2 and B1, between tests B2 and C, between tests A and B1, and between tests A and C. Test B2 resulted in the greatest FB range of motion with a mean (SD) range of 47.5° (0.6°) (95% CI, 45.9-49.2), while test B1 resulted in the least range of motion with a mean (SD) range of 41.3° (0.6°) (95% CI, 39.7-42.9) (Table 2 and Table 3). 
Lumbar Inclinometer (FB at L1)
Controlling for sex, age, height, weight, and current pain (neck/back), statistically significant differences (P<.05) were found between tests B2 and C, between tests A and C, between tests B2 and B1, and between tests A and B1. Test B2 resulted in the greatest FB range of motion with a mean (SD) range of 97.5° (4.4°) (95% CI, 95.1-99.9), while test C resulted in the least range of motion with a mean (SD) range of 83.9° (4.4°) (95% CI, 81.5-86.3) (Table 2 and Table 3). 
Lumbar Region Forward Bending
Controlling for sex, age, height, weight, and current pain (neck/back), statistical differences (P<.05) were found between tests B2 and C, between tests A and C, as well as between tests B2 and B1. Test B2 resulted in the greatest FB range of motion with a mean (SD) range of 50.0° (0.8°) (95% CI, 48.0-51.9) while test C resulted in the least range of motion with a range of 42.6° (0.8°) (95% CI, 40.7-44.5) (Table 2 and Table 3). 
Comment
The results of the present study reject our null hypotheses that cervical spine position does not affect whole trunk FB and lumbar FB. The results confirm, for both whole trunk FB and lumbar FB, our expectations that test B2 (ie, progressive segmental FB, followed by specified cervical BB, followed by whole trunk FB) would result in greater FB range of motion than test B1 (progressive segmental FB) and that test B2 would produce results similar to those of test A (ie, non-specified FB). The results confirm, for whole trunk FB but not for lumbar FB, our expectation that test A would result in greater FB range of motion than test B1 would. We had no expectation regarding test C (ie, specified cervical BB followed by whole trunk FB). These results also demonstrate that FB at the hips did not occur in isolation from bending of the cervical spine, as evidenced by the statistically significant mean differences of sacral inclinometer readings between test B1 and test B2. 
For a clinical test to be meaningful, it must satisfy 3 criteria. First, it must be construct valid; that is, it must potentially measure what it is purported to measure, not something else. When a factor has potential causal impact on the measure, it is a confounding variable. Experiments must control or account for confounding variables, lest the experiment be rendered nongeneralizable and, thus, meaningless. For example, if temperature, pressure, volume, container, and duration influence a chemical reaction, all of which are controlled except for temperature' then the utility of the experimental results must be called into question, as the experiment is unlikely to produce consistent results across locations and experimenters. A clinical test must similarly control for confounding variables. 
Second, a clinical test must be reliable. For the meaningful care of a specific patient by a single caretaker, intraexaminer reliability is requisite to make interpretable comparisons within the patient, as well as about that individual over time. For any purpose potentially involving more than 1 examiner in the care of a specific patient or a population, interexaminer reliability is necessary.54 
Third, a clinical test must have predictive validity. That is, a clinical test must contribute to establishing causal relationships. Predictive validity can be addressed only by controlled experimentation. Recognizing causal factors creates the opportunity for rational exploration of the efficacy of treatments. The present study was not a test of predictive validity; rather, it was a test of construct validity—a necessary prerequisite for meaningful tests of predictive validity. 
Additionally, a test must be safe, practical, and cost effective. Concurrent validity is established when a test produces results similar to those of another valid test. Establishing concurrent validity is often useful for exploring questions of safety, practicality, and cost.41,58,56 
In order for a clinical test of a specific question to be construct valid, it must be capable of answering that specific question. Failure to control for a confounding variable therefore invalidates a test, as the specific question is no longer being addressed. Current AMA protocol (ie, test A) seeks to specifically test lumbar FB by controlling motions at the knees and by subtracting FB at the hips from lumbo-pelvic FB. To further ensure that lumbar FB and whole trunk FB range of motion were specifically tested, tests B1 and B2 were designed to reveal any influence of cervical spine bending on lumbar FB and whole trunk FB range of motion by controlling cervical spine bending direction and sequence. Test B1 specifically asks, “What is the participant's capacity for range of motion during whole trunk FB and lumbar FB while FB sequentially cephalad to caudad the entire spine and pelvis?” Test B2 asks, “What is the participant's capacity for further whole trunk FB and lumbar FB range of motion after progressive segmental FB followed by cervical BB?” 
The statistically significant differences between tests B1 and B2 show that the direction and sequencing of neck bending does indeed influence the total amount of both whole trunk FB and lumbar FB range of motion. Therefore, direction and sequencing of neck bending is a confounding variable when measuring whole trunk FB and lumbar FB range of motion. Current AMA protocol does not control for direction and sequencing of neck bending and, consequently, is invalid to ascertain specific lumbar FB and whole trunk FB range of motion within and between subjects. The statistically significant difference between test B2 and test C further demonstrates that the sequencing of whole spine bending is confounding. 
For the present study to be meaningful, it must be construct valid. Studies have found warm-up,46,56,57 sex,43 age,37,43 height, and time of day42,44,58 to be confounding effects of lumbar motion. The question of whether FB was submaximal because pain inhibition must also be addressed.18 As the present study was a within-subject design, the influence of pain limitation was either constant between tests or, if limiting in 1 test and not another, would only lend credence to the basic conclusion that the direction and sequencing of cervical bending may influence FB, the mechanisms by which remain to be determined. The use of adjusted means accounts for sex, age, height, and weight. Others have controlled for time-of-day effect by waiting at least 2 hours after arising, as the differences in hydration of intervertebral disks are thought to be negligible after 2 hours of weight bearing.58 The current AMA protocol includes a warm-up exercise.46,57,60 The present study did not include warm-up. We believe that our study design, our active study population (ie, students and staff during a normal business day), and the mid-morning through afternoon times of data collection mitigate any distortion due to lack of formal warm-up and time of day. Additionally, test A provided some warm-up for tests B1, B2, and C, and the alternating sequence of tests B and C likely further mitigated any effect of the absence of formal warm-up. Similarly, another potential confound is test/retest phenomena, in that participants might have, as a result of learning or other psychological factors, consciously altered FB behavior if the tests were repeated. So, to ensure that test A was performed in the most unconscious, “everyday” manner possible, we did not measure test/retest phenomena. Further, we believe that our design of 4 repetitions of FB, alternating test C with test B, mitigated the possibility of error from test/retest. 
The present study must also be reliable to be meaningful. The fingertip-to-floor test has been shown to be highly reliable.22 The largest source of error when using inclinometers appears to be the insufficient training of test administrators, resulting in diminished ability to specify bony landmarks and to steadily hold instruments. However, the intraexaminer reliability of trained administrators is considered satisfactory.40,41 Our measuring protocols were clearly demonstrated, practiced, and reviewed prior to the study. In addition, the same examiners tested a given participant (promoting uniform measuring technique within a participant), and the data analysis was within subject. 
Statistically significant differences and clinically meaningful differences are not equivalent concepts. This study was conducted with a largely asymptomatic population—the population from which normative data are gleaned. The first question, then, is whether specifying the direction and sequencing of cervical bending would materially influence currently accepted norms. The mean difference in lumbar FB range of motion between test A and test B1 was less than 2°, and the difference between test A and test B2, less than 3°; neither of these differences was statistically significant at P<.05. In other words, the construct invalid current AMA protocol produced results on average not statistically dissimilar to the construct valid tests B1 and B2. Thus, this study yielded no compelling impetus to renew research into lumbar FB norms. 
The present study does, however, raise the question of clinical significance for an individual patient. Although the mean differences between tests B1 and B2 were marginally meaningful (a bit less than 4°) in more than 15% of the participants, lumbar FB range of motion in test B2 was 7° or greater than in test B1, and in 7 participants it was greater than 14°, with 1 participant demonstrating a 30° difference. Similarly, 15 participants exhibited 15 cm (approximately 7 inches) or more of increased whole trunk FB range of motion in test B2 than in test B1. Thus, for a given participant, the magnitude of difference in either whole trunk FB or lumbar FB as a result of direction and sequencing of the cervical spine may be substantial. These observations raise the question of whether that difference represents inefficient musculoskeletal function (ie, somatic dysfunction), with consequent cervical BB serving as compensatory behavior during FB. The clinical significance of the requirement for compensatory motion may be relative overuse of cervical structures and consequent premature musculoskeletal wear, pain, and eventual damage. This reasoning points to the need for further research discriminating asymptomatic and symptomatic populations—especially patients with chronic axial pain. 
These findings and arguments lend support to core tenets of the osteopathic profession—that the “body is a unit” and that function or dysfunction of 1 area of the body may influence function or dysfunction in “remote” areas.59 The current study examined a subsegment of the body—the spine—and demonstrated that the cervical spine, although historically classified and often clinically viewed as being disjunct from the trunk and lumbar spine, is indeed functionally linked. The possible mechanisms of that linkage can be broadly classified into the following 2 categories: (1) passive contracture or elongation of 1 or more structures (eg, fascia, muscles, tendons, ligaments, disks, dura mater)60-62 and (2) differences in active muscular behavior including tone and patterns of activation. Further research will be necessary to discriminate those possibilities. 
Finally, prior studies6,15,25,26,28,30,31,39,40,57 have shown poor interexaminer reliability for the AMA-sanctioned method for measuring thoracolumbar motions. Repeating those studies while controlling direction and sequencing of cervical bending may reveal improved reliability. Additionally, controlling for cervical bending may help future investigations clarify the relationships between lumbar motion and low back pain. 
Interpretation of these results is limited by an insufficient platform to ensure that no participant could touch the floor during any test, by absence of measured examiner reliability, and by nonspecific characterization of pain. 
Conclusion
Whole trunk FB, lumbar FB, and FB at the hips do not occur in isolation from direction and sequencing of cervical bending, as well as from whole spine bending sequence, highlighting—heretofore unaccounted for—sources of error in lumbar and trunk FB studies. Tests to determine lumbar and trunk FB range of motion must control for direction and sequencing of cervical and whole spine bending to be construct valid. The present study supports the osteopathic tenet that “the body is a unit” and raises several additional questions worthy of further research. It also strongly suggests that, in clinical practice, direction and sequencing of cervical bending, as well as whole spine bending sequence, should be controlled when measuring lumbar and trunk FB range of motion. 
Acknowledgments
We gratefully acknowledge the Nova Southeastern University College of Osteopathic Medicine 2002 Osteopathic Principles and Practice Undergraduate Fellows for their assistance in data collection; Khalil Carter, MD, a visiting fellow of the college, for his assistance in data collection and organization; and John Schousboe, MD, PhD, of Park Nicollet Medical Center in St Louis Park, Minnesota, for his assistance in manuscript preparation. 
   Financial Disclosures: This work was supported by a Wallace Research Foundation grant to the Restorative Care Foundation to Drs Brooks and Patterson and by NSU-COM.
 
References
National Center for Complementary and Alternative Medicine. Spinal manipulation for low back pain. National Center for Complementary and Alternative Medicine Web site. http://nccam.nih.gov/health/pain/spinemanipulation.htm#back-pain. Updated April 2012. Accessed May 22, 2012.
Licciardone JC, Brimhall AK, King LN. Osteopathic manipulative treatment for low back pain: a systematic review and meta-analysis of randomized controlled trials. BMC Musculoskelet Disord. 2005;6:43. http://www.biomedcentral.com/1471-2474/6/43. Accessed May 22, 2012.
Schwab WA. Principles of manipulative treatment—low back problem. In: Barnes MK, ed. 1965 AAO Yearbook. Carmel, CA: Applied Academy of Osteopathy; 1965:90-94.
Kuchera ML. Postural considerations in osteopathic diagnosis and treatment. In: Chila AG, executive ed. Foundations of Osteopathic Medicine. 3rd ed. Baltimore, MD: Lippincott Williams & Wilkins; 2011.
Mayer TG, Tencer AF, Kristoferson S, Mooney V. Use of noninvasive techniques for quantification of spinal range-of-motion in normal subjects and chronic low-back dysfunction patients. Spine (Phila Pa 1976). 1984;9(6):588-595. [CrossRef] [PubMed]
Keeley J, Mayer TG, Cox R, Gatchel RJ, Smith J, Mooney V. Quantification of lumbar function, part 5: reliability of range-of-motion measures in the sagittal plane and an in vivo torso rotation measurement technique. Spine (Phila Pa 1976). 1986;11(1):31-35. [CrossRef] [PubMed]
Triano JJ, Schultz AB. Correlation of objective measure of trunk motion and muscle function with low-back disability ratings. Spine (Phila Pa 1976). 1987;12(6):561-565. [CrossRef] [PubMed]
Burton AK, Tillotson KM, Troup JD. Variation in lumbar sagittal mobility with low-back trouble. Spine (Phila Pa 1976). 1989;14(6):584-590. [CrossRef] [PubMed]
Mellin G. Decreased joint and spinal mobility associated with low back pain in young adults. J Spinal Disord. 1990;3(3):238-243. [PubMed]
Salminen JJ, Maki P, Oksanen A, Pentti J. Spinal mobility and trunk muscle strength in 15-year-old schoolchildren with and without low-back pain. Spine (Phila Pa 1976). 1992;17(4):405-411. [CrossRef] [PubMed]
Cardin AJ, Hadida C. Evaluation of lumbar intersegmental range of motion using flexion-extension radiographs of asymptomatic versus low back pain adults. J Can Chiropr Assoc. 1994;38(2):83-89.
Esola MA, McClure PW, Fitzgerald GK, Siegler S. Analysis of lumbar spine and hip motion during forward bending in subjects with and without a history of low back pain. Spine (Phila Pa 1976). 1996;21(1):71-78. [CrossRef] [PubMed]
Porter JL, Wilkinson A. Lumbar hip flexion motion: a comparative study between asymptomatic and chronic low back pain in 18- to 36-year-old men. Spine (Phila Pa 1976). 1997;22(13s):1508-1513; discussion 1513-1514. [CrossRef] [PubMed]
Snook SH, Webster BS, McGorry RW, Fogleman MT, McCann KB. The reduction of chronic nonspecific low back pain through the control of early morning lumbar flexion: a randomized controlled trial. Spine (Phila Pa 1976). 1998;23(23):2601-2607. [CrossRef] [PubMed]
Barrett CJ, Singer KP, Day R. Assessment of combined movements of the lumbar spine in asymptomatic and low back pain subjects using a three-dimensional electromagnetic tracking system. Man Ther. 1999;4(2):94-99. [CrossRef] [PubMed]
Sullivan MS, Shoaf LD, Riddle DL. The relationship of lumbar flexion to disability in patients with low back pain. Phys Ther. 2000;80(3):240-250. http://ptjournal.apta.org/content/80/3/240.full. Accessed May 22, 2012.
Zuberbier OA, Kozlowski AJ, Hunt DGet al. Analysis of the convergent and discriminant validity of published lumbar flexion, extension, and lateral flexion scores. Spine (Phila Pa 1976). 2001;26(20s):E472-E478. [CrossRef] [PubMed]
Geisser ME, Haig AJ, Wallbom AS, Wiggert EA. Pain-related fear, lumbar flexion, and dynamic EMG among persons with chronic musculoskeletal low back pain. Clin J Pain. 2004;20(2):61-69. [CrossRef] [PubMed]
Wong TK, Lee RY. Effects of low back pain on the relationship between the movements of the lumbar spine and hip. Hum Mov Sci. 2004;23(1):21-34. [CrossRef] [PubMed]
Jones MA, Stratton G, Reilly T, Unnithan VB. Biological risk indicators for recurrent non-specific low back pain in adolescents. Br J Sports Med. 2005;39(3):137-140. [CrossRef] [PubMed]
Hanley EN, Matteri RE, Frymoyer JW. Accurate roentgenographic determination of lumbar flexion-extension. Clin Orthop Relat Res. 1976;115:145-148. [PubMed]
Gauvin MG, Riddle DL, Rothstein JM. Reliability of clinical measurements of forward bending using the modified fingertip-to-floor method. Phys Ther. 1990;70(7):443-447. http://ptjournal.apta.org/content/70/7/443.full.pdf+html. Accessed May 23, 2012.
Mellin G, Kiiski R, Weckström A. Effects of subject position on measurements of flexion, extension, and lateral flexion of the spine. Spine (Phila Pa 1976). 1991;16(9):1108-1110. [CrossRef] [PubMed]
Panjabi M, Chang D, Dvorák J. An analysis of errors in kinematic parameters associated with in vivo functional radiographs. Spine (Phila Pa 1976). 1992;17(2):200-205. [CrossRef] [PubMed]
Rondinelli R, Murphy J, Esler A, Marciano T, Cholmakjian C. Estimation of normal lumbar flexion with surface inclinometry: a comparison of three methods. Am J Phys Med Rehabil. 1992;71(4):219-224. [CrossRef] [PubMed]
Stude DE, Goertz C, Gallinger M. Inter- and intraexaminer reliability of a single, digital inclinometric range of motion measurement technique in the assessment of lumbar range of motion. J Manipulative Physiol Ther. 1994;17(2):83-87. [PubMed]
Yasukouchi A, Isayama T. The relationships between lumbar curves, pelvic tilt and joint mobilities in different sitting postures in young adult males. Appl Human Sci. 1995;14(1):15-21. [PubMed]
Madson TJ, Youdas JW, Suman VJ. Reproducibility of lumbar spine range of motion measurements using the back range of motion device. J Orthop Sports Phys Ther. 1999;29(8):470-477. [CrossRef] [PubMed]
Mannion A, Troke M. A comparison of two motion analysis devices used in the measurement of lumbar spinal mobility. Clin Biomech (Bristol, Avon). 1999;14(9):612-619. [CrossRef] [PubMed]
Nitschke JE, Nattrass CL, Disler PB, Chou MJ, Ooi KT. Reliability of the American Medical Association guides' model for measuring spinal range of motion: its implication for whole person impairment rating. Spine (Phila Pa 1976). 1999;24(3):262-268. [CrossRef] [PubMed]
Tousignant M, Poulin L, Marchand S, Viau A, Place C. The Modified-Modified Schober Test for range of motion assessment of lumbar flexion in patients with low back pain: a study of criterion validity, intra- and inter-rater reliability and minimum metrically detectable change. Disabil Rehabil. 2005;27(10):553-559. [CrossRef] [PubMed]
Rondinelli R. Forward. In: American Medical Association. Guide to the Evaluation of Permanent Impairment. 6th ed. Chicago, IL: American Medical Association; 2008.
Hoppenfeld S. Physical Examination of the Spine and Extremities. New York, NY: Appleton-Century-Crofts; 1976.
Magee D. Orthopedic Physical Assessment. Philadelphia, PA: W.B. Saunders Co; 1984.
Borenstein DG, Wiesel SW, Boden SD. Low Back Pain: Medical Diagnosis and Comprehensive Management. Philadelphia, PA: W.B. Saunders Co; 1995.
Cole AJ, Herring SA. The Low Back Pain Handbook: A Practical Guide for the Primary Care Physician. Philadelphia, PA: Hanley and Belfus, Inc; 1997.
Fitzgerald GK, Wynveen KJ, Rheault W, Rothschild B. Objective assessment with establishment of normal values for lumbar spinal range of motion. Phys Ther. 1983;63(11):1776-1781. http://ptjournal.apta.org/content/63/11/1776.full.pdf+html. Accessed May 23, 2012.
Beattie P, Rothstein JM, Lamb RL. Reliability of the attraction method for measuring lumbar spine backward bending. Phys Ther. 1987;67(3):364-369. http://ptjournal.apta.org/content/67/3/364.full.pdf+html. Accessed May 23, 2012.
Gill K, Krag MH, Johnson GB, Haugh LD, Pope MH. Repeatability of four clinical methods for assessment of lumbar spinal motion. Spine (Phila Pa 1976). 1988;13(1):50-53. [CrossRef] [PubMed]
Mayer RS, Chen IH, Lavender SA, Trafimow JH, Andersson GB. Variance in the measurement of sagittal lumbar spine range of motion among examiners, subjects, and instruments. Spine (Phila Pa 1976). 1995;20(13):1489-1493. [CrossRef] [PubMed]
Mayer TG, Kondraske G, Beals SB, Gatchel RJ. Spinal range of motion: accuracy and sources of error with inclinometric measurement. Spine (Phila Pa 1976). 1997;22(17):1976-1984. [CrossRef] [PubMed]
Wing P, Tsang I, Gagnon F, Susak L, Gagnon R. Diurnal changes in the profile shape and range of motion of the back. Spine (Phila Pa 1976). 1992;17(7):761-766. [CrossRef] [PubMed]
McGregor AH, McCarthy ID, Hughes SP. Motion characteristics of the lumbar spine in the normal population. Spine (Phila Pa 1976). 1995;20(22):2421-2428. [CrossRef] [PubMed]
Ensink FB, Saur PM, Frese K, Seeger D, Hildebrandt J. Lumbar range of motion: influence of time of day and individual factors on measurements. Spine (Phila Pa 1976). 1996;21(11):1339-1343. [CrossRef] [PubMed]
Rice J, Kaliszer M, Walsh M, Jenkinson A, O'Brien T. Kinematics of the toe touching test: an investigation using motion analysis. Clin Anat. 2004;17(2):130-138. [CrossRef] [PubMed]
Gerhardt JJ, Cocchiarella L, Lea RD. The Practical Guide to Range of Motion Assessment. Chicago, IL: American Medical Association; 2002.
Stokes IA, Abery JM. Influence of the hamstring muscles on lumbar spine curvature in sitting. Spine (Phila Pa 1976). 1980;5(6):525-528. [CrossRef] [PubMed]
Bridger RS, Wilkinson D, van Houweninge T. Hip joint mobility and spinal angles in standing and in different sitting postures. Hum Factors. 1989;31(2):229-241. [PubMed]
Tsuji T, Matsuyama Y, Goto Met al. Knee-spine syndrome: correlation between sacral inclination and patellofemoral joint pain. J Orthop Sci. 2002;7(5):519-523. [CrossRef] [PubMed]
Shin G, Shu Y, Li Z, Jiang Z, Mirka G. Influence of knee angle and individual flexibility on the flexion-relaxation response of the low back musculature. J Electromyogr Kinesiol. 2004;14(4):485-494. [CrossRef] [PubMed]
Rogers FJ, D'Alonzo GEJr, Glover JCet al. Proposed tenets of osteopathic medicine and principles for patient care. J Am Osteopath Assoc. 2002;102:63-65. [PubMed]
Lew PC, Morrow CJ, Lew AM. The effect of neck and leg flexion and their sequence on the lumbar spinal cord: implications in low back pain and sciatica. Spine (Phila Pa 1976). 1994;19(21):2421-2424; discussion 2425. [CrossRef] [PubMed]
Rice J, Walsh M, Jenkinson A, O'Brien TM. Measuring movement at the low back. Clin Anat. 2002;15(2):88-92. [CrossRef] [PubMed]
Lang T, Secic M. How to Report Statistics in Medicine. 2nd ed. Philadelphia, Pennsylvania: American College of Physicians; 2006.
Gajdosik RL, Bohannon RW. Clinical measurement of range of motion: review of goniometry emphasizing reliability and validity. Phys Ther. 1987;67(12):1867-1872. http://ptjournal.apta.org/content/67/12/1867.full.pdf+html. Accessed May 23, 2012.
Atha J, Wheatley DW. The mobilising effects of repeated measurement on hip flexion. Br J Sports Med. 1976;10:22-24. [CrossRef] [PubMed]
Frost M, Stuckey S, Smalley LA, Dorman G. Reliability of measuring trunk motions in centimeters. Phys Ther. 1982;62(10):1431-1437. http://ptjournal.apta.org/content/62/10/1431.full.pdf+html. Accessed May 23, 2012.
Adams MA, Dolan P, Hutton WC. Diurnal variations in the stresses on the lumbar spine. Spine (Phila Pa 1976). 1987;12(2):130-137. [CrossRef] [PubMed]
Tenets of osteopathic medicine. American Osteopathic Association Web site. http://www.osteopathic.org/inside-aoa/about/leadership/Pages/tenets-of-osteopathic-medicine.aspx. Accessed June 18, 2012.
O'Connell JEA. Clinical signs of meningeal irritation. Brain. 1946;69:9-21. [CrossRef]
Breig A, Troup JDG. Biomechanical considerations in the straight-leg-raising test: cadaveric and clinical studies of the effects of medial hip rotation. Spine (Phila Pa 1976). 1979;4(3):242-250. [CrossRef] [PubMed]
LaBan MM, Macy JA, Meerschaert JR. Intermittent cervical traction: a progenitor of lumbar radicular pain. Arch Phys Med Rehabil. 1992;73(3):295-296. [PubMed]
Table 1.
Descriptive Statistics of Participants in a Study on Cervical Spine Bending and Forward Bending
Variable Mean (SD) Range
Age, ya 33.4 (12.6) 21.0-75.0
Height, inb 67.2 (4.4) 59.0-83.0
Weight, lbc 162.6(35.5) 90.0-260.0
Sexd
Male 123 (52.6) NA
Female 111 (47.4) NA
Paind
Y 59 (25.2) NA
N 175 (74.8) NA
  a n=236
  b n=235
  c n=236
  d Data presented as No. (%)
  Abbreviations: NA, not applicable; SD, standard deviation
Table 1.
Descriptive Statistics of Participants in a Study on Cervical Spine Bending and Forward Bending
Variable Mean (SD) Range
Age, ya 33.4 (12.6) 21.0-75.0
Height, inb 67.2 (4.4) 59.0-83.0
Weight, lbc 162.6(35.5) 90.0-260.0
Sexd
Male 123 (52.6) NA
Female 111 (47.4) NA
Paind
Y 59 (25.2) NA
N 175 (74.8) NA
  a n=236
  b n=235
  c n=236
  d Data presented as No. (%)
  Abbreviations: NA, not applicable; SD, standard deviation
×
Table 2.
Forward Bending Range of Motion With Cervical Spine Bending (N=232)
Test Mean (SD) FB Range of Motion 95% CI
Whole-Trunk FB, cm
A 36.63 (2.57) 35.32-37.94
B1 41.45 (2.66) 40.14-42.76
B2 34.43 (2.65) 33.13-35.74
C 50.18 (2.58) 48.87-51.48
Sacrum FB,°
A 46.9 (0.7) 45.3-48.6
B1 41.3 (0.6) 39.7-42.9
B2 47.5 (0.6) 45.9-49.2
C 41.3 (0.6) 39.7-43.0
L1 FB,°
A 94.4 (4.0) 92.1-96.8
B1 87.3 (4.2) 84.9-89.7
B2 97.5 (4.4) 95.1-99.9
C 83.9 (4.4) 81.5-86.3
Lumbar FB,°
A 47.5 (0.9) 45.6-49.4
B1 46.0 (0.8) 44.1-47.9
B2 50.0 (0.8) 48.0-51.9
C 42.6 (0.8) 40.7-44.5
  Abbreviations: CI, confidence interval; FB, forward bending; SD, standard deviation.
Table 2.
Forward Bending Range of Motion With Cervical Spine Bending (N=232)
Test Mean (SD) FB Range of Motion 95% CI
Whole-Trunk FB, cm
A 36.63 (2.57) 35.32-37.94
B1 41.45 (2.66) 40.14-42.76
B2 34.43 (2.65) 33.13-35.74
C 50.18 (2.58) 48.87-51.48
Sacrum FB,°
A 46.9 (0.7) 45.3-48.6
B1 41.3 (0.6) 39.7-42.9
B2 47.5 (0.6) 45.9-49.2
C 41.3 (0.6) 39.7-43.0
L1 FB,°
A 94.4 (4.0) 92.1-96.8
B1 87.3 (4.2) 84.9-89.7
B2 97.5 (4.4) 95.1-99.9
C 83.9 (4.4) 81.5-86.3
Lumbar FB,°
A 47.5 (0.9) 45.6-49.4
B1 46.0 (0.8) 44.1-47.9
B2 50.0 (0.8) 48.0-51.9
C 42.6 (0.8) 40.7-44.5
  Abbreviations: CI, confidence interval; FB, forward bending; SD, standard deviation.
×
Table 3.
Cervical Spine Bending and Forward Bending (FB): Generalized Estimating Equations Model With Covariates (N=232)
Test Mean Difference Between Tests (95% Confidence Interval)
Lumber FB Test B1 B2 C
A 1.5 (-2.1-5.1) 2.5 (-1.1-6.0) 4.93 (1.4-8.5)a
B1 4.0 (0.4-7.5)a 3.4 (-0.1-7.0)
B2 7.4 (3.8-11)a
Sacrum FB Test
A 5.7 (2.8-8.6)a 0.5 (-2.4-3.4) 5.5 (2.6-8.4)a
B1 6.2 (3.3-9.1)a 0.1 (-2.8-3.0)
B2 6.1 (3.2-9.0)a
L1 FB
A 7.1 (2.7-11.6)a 3.1 (-1.4-7.5) 10.5 (6.1-15.00)a
B1 -10.2 (5.8-14.6)a 3.4 (-1.0-7.8)
B2 13.6 (9.2-18.0)a
Whole-Trunk FB
A 4.82 (2.39-7.26)a 2.13 (-0.30-4.57) 13.59 (11.15-16.03)a
B1 6.96 (4.52-9.40)a 8.76 (6.33-11.20)a
B2 15.72 (13.29-18.16)a
  a Difference was statistically significant (P<.05).
Table 3.
Cervical Spine Bending and Forward Bending (FB): Generalized Estimating Equations Model With Covariates (N=232)
Test Mean Difference Between Tests (95% Confidence Interval)
Lumber FB Test B1 B2 C
A 1.5 (-2.1-5.1) 2.5 (-1.1-6.0) 4.93 (1.4-8.5)a
B1 4.0 (0.4-7.5)a 3.4 (-0.1-7.0)
B2 7.4 (3.8-11)a
Sacrum FB Test
A 5.7 (2.8-8.6)a 0.5 (-2.4-3.4) 5.5 (2.6-8.4)a
B1 6.2 (3.3-9.1)a 0.1 (-2.8-3.0)
B2 6.1 (3.2-9.0)a
L1 FB
A 7.1 (2.7-11.6)a 3.1 (-1.4-7.5) 10.5 (6.1-15.00)a
B1 -10.2 (5.8-14.6)a 3.4 (-1.0-7.8)
B2 13.6 (9.2-18.0)a
Whole-Trunk FB
A 4.82 (2.39-7.26)a 2.13 (-0.30-4.57) 13.59 (11.15-16.03)a
B1 6.96 (4.52-9.40)a 8.76 (6.33-11.20)a
B2 15.72 (13.29-18.16)a
  a Difference was statistically significant (P<.05).
×