[ research report ]

Brad D. Campbell, PT1 Suzanne J. Snodgrass, PT, PhD2

The Effects of Thoracic Manipulation

on Posteroanterior Spinal Stiffness

S
pinal manipulation has been used for hundreds of years40 cervical,4,9,36 thoracic,39 and lum-
bar spine pain when clinically
and is commonly performed by physical therapists,
indicated.4,7,10 Mechanisms to ex-
osteopaths, chiropractors, and medical practitioners.26
Downloaded from www.jospt.org at on December 7, 2015. For personal use only. No other uses without permission.

plain how spinal manipulation

Spinal manipulation is defined as a small amplitude works are not well defined and
manipulative thrust technique performed with speed.35 there is a lack of evidence to sup-
Manipulations to the thoracic spine are typically applied so that port existing theories.2
Therapists usually apply ma-
the emphasis of movement is localized to a single intervertebral
SUPPLEMENTAL nipulation to spinal levels that
Copyright 2010 Journal of Orthopaedic & Sports Physical Therapy. All rights reserved.

level to restore its range of movement.26,35 Spinal manipulation has been
shown to be an effective form of treatment for patients with
TTSTUDY DESIGN: Controlled laboratory study,
with measurements taken before and after a
standardized clinical intervention
TTOBJECTIVES: To determine if thoracic ma-
nipulation alters the posteroanterior (PA) spinal
stiffness of the thoracic spine, and the factors
Journal of Orthopaedic & Sports Physical Therapy
VIDEO ONLINE
following manipulation. Generalized linear mixed-
models were used to determine if subject age,
gender, spinal level, premanipulation stiffness, or
manipulative thrust parameters were associated
with postmanipulation stiffness.
TTRESULTS: Thoracic spine PA stiffness differed
they assess as being stiffer or more
symptomatic than others, direct-
ing their treatment toward generating
potential mechanical effects where they
are most likely to produce the best clinical
outcome for a patient.35 This is supported
by some evidence for an association be-
tween increased spinal stiffness and pain-
ful symptoms in both the lumbar30 and

associated with any potential changes in stiffness.
TTBACKGROUND: Spinal manipulation is com-
monly used to treat thoracic pain and dysfunction.
Therapists use manual assessment of PA spinal
stiffness to determine the appropriateness and ef-
fectiveness of treatment, with potential changes in
spinal stiffness possibly contributing to symptom-
atic improvement following manipulation.
TTMETHODS: Thoracic PA spinal stiffness was
measured at 5 vertebral levels (manipulated level
and 2 levels above and below), in 24 asymptom-
atic subjects, before and after manipulation. Five
cycles of standardized mechanical PA force were
applied to the spinous process while recording
resistance to movement and concurrent displace-
ment, with stiffness defined as the slope of the
linear portion of the force-displacement curve.
A 2-way repeated-measures analysis of variance
determined differences between premanipulation
and postmanipulation among multiple spinal
levels. Linear regression determined the relation-
ship between stiffness magnitude and its change
between spinal levels (F4,92 = 21.1, P<.001) but was
not significantly different following manipulation.
The mean change in spinal stiffness correlated
with stiffness magnitude at the manipulated spinal
level only but not other levels (Pearson r, 0.65;
P<.001). Greater postmanipulation stiffness was
associated with being male (regression coefficient,
1.16; 95% CI: 0.52, 1.79; P<.001) and with higher
premanipulation stiffness (regression coefficient,
0.63; 95% CI: 0.49, 0.77; P<.001). Manipulation
force parameters were not associated with post-
manipulation stiffness.
TTCONCLUSION: In asymptomatic individuals,
thoracic PA spinal stiffness is not significantly
different when measured before and after thrust
manipulation, but any potential mechanical effects
appear associated with the manipulated spinal
level rather than other levels. J Orthop Sports
Phys Ther 2010;40(11):685-693. doi:10.2519/
jospt.2010.3271
TTKEY WORDS: physical therapy techniques,
spinal manipulation, thoracic vertebrae
cervical spines.47 Intra-articular mechani-
cal effects of thoracic spine manipulation
have been hypothesized to restore normal
movement of the intervertebral joints
by altering articular adhesions,16,46 sug-
gested by increased zygapophyseal joint
gapping on magnetic resonance imaging
following spinal manipulation.12 It has
also been suggested that thoracic spine
manipulation restores localized biome-
chanics by reducing the resistive forces
of surrounding vertebral segments,15 po-
tentially producing range-of-movement
changes in the whole spine.17 Therefore, a
clinical aim of manipulation is to attempt
to reduce palpable intervertebral joint
stiffness hypothesized to be the cause of
a patients symptoms.35 However, the me-
chanical effects of manipulation have not
been documented.16,25

1
Staff Physiotherapist, Calvary Mater Hospital, Newcastle, NSW, Australia. 2 Senior Lecturer, Discipline of Physiotherapy, School of Health Sciences, The University of Newcastle,
Newcastle, NSW, Australia. This research was approved by The Institutional Review Board of the University of Newcastle (The Human Research Ethics Committee). Address
correspondence to Dr Suzanne Snodgrass, Discipline of Physiotherapy, School of Health Sciences, The University of Newcastle, Callaghan, NSW 2308, Australia. E-mail: Suzanne.
Snodgrass@newcastle.edu.au

journal of orthopaedic & sports physical therapy|volume 40|number 11|november 2010|685

40-11 Campbell.indd 685 10/20/10 1:00 PM

 [ research report
Clinically, spinal posteroanterior
(PA) stiffness is described as the per-
ceived resistance through intervertebral
joint range of movement during the ap-
plication of manual forces applied by
therapists.14,35 Therapist assessment of
differences in the amount, behavior, and
quality of PA stiffness between adjacent
vertebral levels is used to identify symp-
tomatic joints14 that may require manip-
ulation treatment.35 Therapists may also
palpate spinal stiffness when assessing
responses to treatment.14,26 Further, in-
Downloaded from www.jospt.org at on December 7, 2015. For personal use only. No other uses without permission.
creased therapist-assessed spinal stiff-
ness has been associated with symptoms,
with 71% of 131 patients with low back
pain in 1 study judged to have hypomobil-
ity.19 Objectively, PA spinal stiffness has
been measured by mechanical devices
Copyright 2010 Journal of Orthopaedic & Sports Physical Therapy. All rights reserved.
applying standardized forces and defined
as the relationship between the tissues
resistance to the force applied and the
concurrent displacement.41 An increase
in mechanically assessed spinal stiffness
has also been associated with low back
symptoms.30
Mechanical changes in PA spinal
stiffness following spinal manipulation
Journal of Orthopaedic & Sports Physical Therapy
may be affected by the properties of the
manual forces that are applied. Manipu-
lation force parameters that have been
measured include preload force,27 peak
force,18,22,27 thrust duration,18,22,27 thrust
rate,11,18,22 and the production of cavita-
tion.27 Cavitation describes the audible
crack heard during manipulation, which
is hypothesized to occur as the joint
surfaces separate, creating a gas bubble
which is collapsed by the surrounding
fluid.3 The possible effect of manipulation
forces and cavitation on the mechanical
behavior of spinal joints has not been
investigated.
This study investigates the mechani-
cal effects of thoracic manipulation. The
aim is to determine if thoracic PA spinal
stiffness changes following manipula-
tion, and whether there are specific fac-
tors, such as subject age, gender, baseline
spinal stiffness, and applied manipula-
tion force parameters that are associated
with the magnitude of potential changes
686|november 2010|volume 40|number 11|journal of orthopaedic & sports physical therapy
40-11 Campbell.indd 686
in stiffness following manipulation.
METHODS
Equipment and Measurement
Spinal Stiffness Thoracic spine stiffness
was measured before and after manipu-
lation, using a custom-designed stiffness
assessment device.45 The device applies 5
repetitions of standardized force at 1 Hz
to the spinous process via an oscillating
steel rod. It measures the displacement
of the rod in millimeters, using a linear
variable differential transformer (DC-EC
1000; Schaevitz Sensors, Lucas Control
Systems, Hampton, VA), while concur-
rently measuring the tissues resistance to
movement in Newtons using a load cell
(UMM-K050; Dacell Co, Ltd, Chungbuk,
South Korea). The relationship between
this resistance and the simultaneous
displacement represents the spinal stiff-
ness. Stiffness is defined as the average
slope of the linear region of the force-dis-
placement curves for oscillation cycles 2
through 5.5,6,29,33,49 The first cycle is usual-
ly discarded, as it is significantly different
to the subsequent 4 cycles.29 The linear
region was standardized for all measure-
ments as between 30 to 100 N of force,
as in previous studies,29,41,42,49 because PA
spinal stiffness has been shown to differ
when calculated using different levels of
applied force.31 The data from the linear
variable differential transformer and load
cell were collected and processed using
LabVIEW 8.0. The stiffness assessment
device has adequate reliability (intraclass
correlation coefficient [ICC2,1] = 0.99;
95% confidence interval [CI]: 0.93, 1.00,
when tested with inert materials of vary-
ing stiffness; ICC = 0.84; 95% CI: 0.74,
0.90, for human stiffness measurement
performed on the cervical spine).45 The
standard error of measurement for the
device is 0.83 N/mm for measurements
performed at C7. Data are not available
for the thoracic spine, but it is expected
that reliability would be similar, as an
identical protocol was used, and the de-
vice is designed to test a wide range of
stiffness levels.
]
FIGURE 1. Photograph of the stiffness assessment
device positioned on a subject at vertebral level T7
with a sagittal angle of 2 in the caudad direction.
Spinal stiffness measurements were
performed while the subject lay prone
on a standard treatment surface (SX3
Physioline Series, Model No 50251; Chat-
tanooga Group, Inc, Sydney, Australia),
positioning the steel rod perpendicular to
the spinal curvature similar to a therapist
performing a PA spinal stiffness assess-
ment clinically. The angle of the device
was standardized for each spinal level
based on the average curvature of the
thoracic spine from a radiographic study
of normal asymptomatic individuals,20 as
previous research indicates applied force
angle affects spinal stiffness measure-
ments.5 The angle was measured using
a 180 protractor built into the device.
Positioning the steel rod onto the subject
was standardized by moving it toward
the skin and immediately stopping when
a slight indentation became visible45
(FIGURE 1, ONLINE VIDEO). Subjects were in-
structed to hold their breath at the end
of exhalation (functional residual capac-
ity) while the measurement was taken,
because spinal stiffness has been shown
to increase on inspiration.43
Manipulation Force Thoracic manipu-
lation forces were measured using an
instrumented treatment table fitted with
10/20/10 1:00 PM

400
300
200
Force (N)
Downloaded from www.jospt.org at on December 7, 2015. For personal use only. No other uses without permission.
100
0 and the subject positioned on the table
Copyright 2010 Journal of Orthopaedic & Sports Physical Therapy. All rights reserved.
A prior to the thrust.
100
1 2
Journal of Orthopaedic & Sports Physical Therapy
FIGURE 2. Typical thoracic manipulation force-time history. (A) Baseline, (B) preload force maximum, (C) preload
force minimum, (D) peak force, (E) total thrust force, (F) thrust duration.
7 load cells able to measure forces in 3
planes.44 In this study, vertical forces cap-
tured at 100 Hz from 4 load cells (X-tran
S1W; Applied Measurement Australia
Pty Ltd, Sydney, Australia) represented
the manipulation forces, as the manipu-
lation technique was applied primarily in
the vertical direction. A previous study
of manipulation forces only reported the
vertical direction of force,13 and another
study suggested vertical force as being
the major component, with horizon-
tal shear forces being relatively small.27
Manipulation force data from the load
cells were collected using PowerLab
data acquisition and Chart Version 4.2.4
software (ADInstruments, Castle Hill,
Australia). Accuracy and reliability of the
instrumented table have been previously
reported as high for vertical forces, with
a mean SD absolute error of 1.1 1.5 N,
journal of orthopaedic & sports physical therapy|volume 40|number 11|november 2010|687
40-11 Campbell.indd 687
D
E
B the thrust technique (FIGURE 2). This al-
C Participants
F
3 4
Time (s)
when tested using known weights.44
Manipulation force-time history
graphs were used to determine data
points for manipulation thrust analysis.
Five biomechanical parameters were
identified: preload force maximum (N),
preload force minimum (N), peak force
(N), thrust duration (seconds), and thrust
rate (N/second) (FIGURE 2). Preload force
maximum is defined as the highest force
during the 500 milliseconds prior to the
beginning of the thrust phase and rep-
resents the therapist moving the spinal
level to the end of normal range of move-
ment.11 Preload force minimum is the
minimum force immediately preceding
the manipulative thrust and represents
the release of force prior to thrust.27 Peak
force is the maximum force produced
during the manipulative thrust.18,22,27
Thrust duration is the time from pre-
load force minimum to peak force.18,22,27
Thrust rate is the difference between
peak force and preload force minimum
divided by the thrust duration.11,18,22 Total
thrust force (N) was calculated to quan-
tify the amount of force that was applied
during the thrust phase of the manipula-
tion and is defined as the difference be-
tween preload force minimum and peak
force. The starting point (baseline zero)
for the identification of these data points
was defined as the point of lowest force
preceding the thrust technique, because
it was identifiable as the period where
the subject expired immediately prior to
lowed the baseline point to include the
combined body weight of the therapist
Asymptomatic participants between 18
and 45 years of age were recruited by
5 placing flyers around the University of
Newcastle (Australia) campus where
data were collected. Subjects were re-
stricted to this age range to avoid po-
tential damage to epiphyseal growth
plates in the young, and in older sub-
jects to reduce the possibility of having
undiagnosed conditions that might con-
traindicate the application of thoracic
manipulation or increase complication
risks. This age range is also similar to
that used in a previous study of the ef-
fects of thoracic manipulation.32 Subjects
were excluded if they had sought treat-
ment for thoracic spine pain in the past
12 months, were pregnant or breast feed-
ing, had a previous spinal surgery, or had
contraindications to thoracic manipula-
tion that included osteoporosis, osteomy-
elitis, rheumatoid arthritis, malignancy,
ankylosing spondylitis, gross joint laxity,
spondylolisthesis, vertebral disc hernia-
tion, diseases of the spinal cord, or cauda
equina syndrome.35 Ethical approval was
provided by The University of Newcastle
Human Research Ethics Committee. All
subjects gave informed consent and their
rights were protected.
10/20/10 1:00 PM
 [ research report ]

TABLE 1 Force-Time History Parameter Data (n = 24)*

FIGURE 3. Thoracic manipulation technique
performed on the instrumented treatment table.
Downloaded from www.jospt.org at on December 7, 2015. For personal use only. No other uses without permission.
Measure Outcome
Preload force maximum (N) 43 23
Preload force minimum (N) 41 27
Peak force (N) 301 69
Total thrust force (N) 342 79
Thrust duration (s) 0.12 0.01
Thrust rate (N/s) 2929 792
* Data presented as mean SD.

Procedure and Data Collection
Vertebral Level Selection An experi-
enced physical therapist researcher pre-
marked the thoracic spinous processes
using standardized clinical methods.21,24
The spinous process of C7 was identified
Copyright 2010 Journal of Orthopaedic & Sports Physical Therapy. All rights reserved.
reproducible behavior of the soft tissues,
as stiffness measurements are usually re-
peatable after preconditioning with only
small differences.6,32,33 There was ap-
proximately 5 minutes between the first
and second (premanipulation) stiffness
had any adverse events with the manipu-
lation technique.
Data Analysis
As the main effect of manipulation was
expected to occur at the spinal level that

by counting the spinous processes cau-
dally from C2 and by verifying the level
using a commonly recommended cervi-
cal extension motion test.38 The thoracic
spinous processes were then marked by
counting down each successive verte-
bral level to T12. An independent expert
physical therapist (not one of the re-
searchers), with postgraduate training in
Journal of Orthopaedic & Sports Physical Therapy
measurements. Stiffness was measured
again at the same 5 levels in randomized
order following manipulation.
Thoracic Manipulation The manipula-
tion procedure consisted of a standard
thoracic technique, as described by Mai-
tland et al,35 involving a PA thrust ap-
plied to the subject in supine lying on
the instrumented treatment table. The
was manipulated, an initial comparison
of PA stiffness measured at the manipu-
lated level before and after manipula-
tion was performed using a paired t test.
To examine possible manipulation ef-
fects on different spinal levels, a 2-way
repeated-measures analysis of variance
(ANOVA) was used. The factors included
in the ANOVA were time (before or af-

manipulative therapy and 20 years clini-
cal experience, then performed standard
clinical assessment to select the thoracic
vertebral level perceived to be the stiffest.
The therapist was instructed to perform
their usual clinical assessment to iden-
tify the stiffest level, and they primarily
used PA passive accessory intervertebral
movements35 applied with the heel of the
hand with the subject positioned prone.
Five spinous processes were labeled for
stiffness measurement. These included
the stiffest spinal level as clinically per-
ceived by the therapist and the 2 levels
cephalad and caudad of it. Manipulation
was performed at the therapist-selected
stiffest level.
Stiffness Measurement Stiffness mea-
surement of the 5 spinal levels was per-
formed in random order and measured
twice prior to the manipulation. The first
measurement of each spinal level was
performed to precondition the tissues.
Preconditioning is used to achieve more
manipulation procedure involved the
subjects linking their hands behind their
neck while the therapist positioned the
spinous process of the selected vertebral
level between the base of their thumb
and middle finger. The therapists thrust-
ing hand and forearm applied a high-ve-
locity, low-amplitude downward thrust
through the elbows of the subject (FIG-
URE 3, ONLINE VIDEO). The subjects age and
gender, the number of thrust attempts,
and the production of cavitation were
also recorded. Cavitation was recorded
if the therapist performing the manipu-
lation heard or felt a cracking sound dur-
ing the thrust technique.23 The therapist
performed thoracic manipulation as it
would be done clinically; so, if cavita-
tion did not occur on the initial thrust,
the technique was repeated until either
cavitation was elicited or the therapist
determined that a localized stretch had
been satisfactorily achieved ( judging by
the compliance of tissues). No subjects
ter manipulation) and spinal level (the
manipulated level, plus 1 and 2 levels
above and below the manipulated level).
To examine whether a potential change
in stiffness after manipulation was as-
sociated with the magnitude of stiffness,
a regression analysis was performed
between the individuals change in stiff-
ness and the average of the individuals
premanipulation and postmanipulation
stiffness values.
Generalized linear mixed-models
were used to examine factors potentially
related to spinal stiffness following ma-
nipulation, adjusting for premanipula-
tion stiffness. The predictor variables
included in the model were subject age,
gender (coded as female = 0, male = 1),
spinal level, number of thrust attempts,
and the following data from the final
thrust attempt: preload force maxi-
mum, preload force minimum, peak
force, total thrust force, thrust dura-
tion, thrust rate, and the production of

688|november 2010|volume 40|number 11|journal of orthopaedic & sports physical therapy

40-11 Campbell.indd 688 10/20/10 1:00 PM

 icantly different following manipulation.
8
There was a significant effect for spinal
level (F4,96 = 8.69, P<.001), indicating
6 that PA stiffness was significantly differ-
ent for different spinal levels relative to
their proximity to the manipulated level.
4 There was no interaction effect between
time and level (P = .61), indicating that
there were no differences in change in
2
Change in Stiffness (N/mm)

spinal stiffness (from premanipulation to

postmanipulation) between the different
0
spinal levels. The mean change in spinal
stiffness was correlated with spinal stiff-
Downloaded from www.jospt.org at on December 7, 2015. For personal use only. No other uses without permission.

ness values at the manipulated spinal lev-

2 el only (Pearson r = 0.65; P<.001), and
not at other levels. At the manipulated
spinal level, when stiffness was high pre-
4
manipulation, it tended to decrease after
manipulation (FIGURE 4).The variables as-
Copyright 2010 Journal of Orthopaedic & Sports Physical Therapy. All rights reserved.

6
sociated with postmanipulation stiffness
were premanipulation stiffness, gender,
and spinal level. Greater postmanipula-
8 tion stiffness was associated with being
male (regression coefficient, 1.16; 95% CI:
0 5 10 15 20 25
0.52, 1.79; P<.001) and with higher pre-
manipulation stiffness (regression coeffi-
Premanipulation Stiffness (N/mm) cient, 0.63; 95% CI: 0.49, 0.77; P<.001).
For the gender variable, this indicates
Journal of Orthopaedic & Sports Physical Therapy

Male Female
FIGURE 4. Relationship between thoracic posteroanterior spinal stiffness and its change at the manipulated spinal
level following manipulation. Negative change represents a decrease in stiffness (Pearson r = 0.65; P<.001).
cavitation. As there were a large number participated in this study. The thoracic
of variables potentially associated with spine level manipulated by the therapist
stiffness, each variable was first exam- was from T3 to T10 for individual sub-
ined in a univariate analysis, and those jects. Between 1 and 5 thrust attempts
variables with a P value of less than .25 were recorded for each subject. Cavita-
in the univariate analysis were entered tion occurred in 13 subjects (54%). Ma-
into a final model.37 To account for vari- nipulation force parameter values are
ability among subjects, premanipula- presented in TABLE 1.
tion stiffness at the manipulated level PA spinal stiffness data for the ma-
was included in each univariate model nipulated level are presented in TABLE 2.
and in the final model. All analyses were There was no change in PA spinal stiff-
performed in STATA 11.0 (StataCorp, ness at the manipulated level following
College Station, TX). manipulation (mean difference between
premanipulation and postmanipulation
RESULTS stiffness, 0.55 N/mm; 95% CI: 1.54,
0.44; P = .26). The 2-way ANOVA dem-
Twenty-four asymptomatic subjects onstrated no effect for time (F1,96 = 0.22, P
(14 males and 10 females; mean SD = .64), indicating that PA stiffness across
age, 21.9 2.5 years; range, 19-31 years) the measured spinal levels was not signif-
that males had 0.16 N/mm greater post-
manipulation stiffness on average than
females, adjusted for premanipulation
stiffness and spinal level. As spinal level
was a categorical variable, it remained in
the model with P<.001 (based on the like-
lihood ratio statistic), indicating that PA
stiffness is significantly different at the
different spinal levels measured.
DISCUSSION
This study investigated the mechani-
cal effects of manipulation. The results
suggest that there is no statistically
significant difference between prema-
nipulation and postmanipulation spi-
nal stiffness in the thoracic spine in
asymptomatic individuals. However,
the mechanical effects of manipulation
appear to occur primarily at the spinal
level that is targeted for the manipula-
tion technique. There was an association
between stiffness and its change (from

journal of orthopaedic & sports physical therapy|volume 40|number 11|november 2010|689

40-11 Campbell.indd 689 10/20/10 1:00 PM

 [ research report ]
Thoracic Posteroanterior Spinal Stiffness Data (N/mm)
TABLE 2
at the Manipulated Level (n = 24)

Subject Number Level Manipulated Cavitation Initial Stiffness (Preconditioning) Premanipulation Stiffness Postmanipulation Stiffness
1 T4 12.69 11.43 12.22
2 T10 17.90 17.22 17.44
3 T4 10.94 11.98 10.94
4 T6 14.78 13.80 13.00
5 T4 10.89 14.36 14.26
6 T6 19.91 18.39 17.08
7 T6 18.82 20.89 17.28
8 T4 11.59 6.17 11.80
9 T9
Downloaded from www.jospt.org at on December 7, 2015. For personal use only. No other uses without permission.

9.24 10.49 12.42

10 T4 9.92 13.47 13.25
11 T4 13.48 13.26 13.08
12 T3 10.52 7.76 9.85
13 T4 18.64 22.82 16.49
14 T5 13.51 15.32 11.95
Copyright 2010 Journal of Orthopaedic & Sports Physical Therapy. All rights reserved.

15 T3 17.46 16.61 17.17

16 T4 11.80 12.81 12.34
17 T4 11.51 12.09 13.55
18 T4 19.03 17.13 16.06
19 T5 10.65 12.79 10.94
20 T4 14.67 14.71 16.21
21 T5 11.64 11.97 11.39
22 T3 22.95 19.59 16.67
23 T5 15.78 15.09 12.60
24 T5
Journal of Orthopaedic & Sports Physical Therapy

15.40 14.92 13.86

14.32 3.74* 14.39 3.84* 13.83 2.36*
* Data presented as mean SD.

premanipulation to postmanipulation)
at the targeted level, but this association
did not occur at the adjacent spinal lev-
els measured. Greater postmanipulation
stiffness was associated with being male
or having a higher level of spinal stiffness.
Manipulation force parameters were not
associated with postmanipulation spinal
stiffness, suggesting that the amount of
force and how it is applied are not related
to reducing stiffness.
The observed differences between
premanipulation and postmanipulation
stiffness for individuals following ma-
nipulation were probably large enough
to be detectable by therapists. The mag-
nitude of differences in stiffness for in-
dividuals following manipulation ranged
between 0.11 to 6.33 N/mm (absolute
value of the change). Physical therapists
can detect a difference in stiffness of ap-
proximately 0.8 N/mm when compar-
ing stiffness stimuli in the range of 6 to
11 N/mm.34 The majority of subjects (17
of 24) in the current study had a differ-
ence between premanipulation and post-
manipulation stiffness of this magnitude
or greater, suggesting that the observed
differences would potentially be detect-
able by therapists. This suggests that the
manual assessment of stiffness following
manipulation may be appropriate in the
clinical setting.
In examining the data for possible
reasons for the nonsignificant change in
stiffness following manipulation, it was
noted that there were 2 subjects with
very low premanipulation stiffness val-
ues (6.17 and 7.76 N/mm compared to
the mean of 14.39 N/mm). These values
were 2.14 and 1.73 SD from the mean,
respectively. It was hypothesized that
subjects who were initially stiffer would
be more likely to demonstrate a decrease
in stiffness following manipulation.
When these 2 subjects are excluded from
analysis, there is a significant mean de-
crease in spinal stiffness at the manipu-
lated level following manipulation in the
remaining 22 subjects (mean change,
0.95 N/mm; 95% CI: 1.81, 0.10; P
= .031). This suggests that manipula-
tion might have a potentially beneficial
mechanical effect on stiffer spines. How-
ever, a larger study focusing on subjects
with stiffer spines would be needed to
confirm this.

690|november 2010|volume 40|number 11|journal of orthopaedic & sports physical therapy

40-11 Campbell.indd 690 10/20/10 1:00 PM

 Stiffness Variations Between Vertebral
Levels
Stiffness differed between thoracic spine
levels in the current study. Mean prema-
nipulation stiffness at T5 was 15.1 N/mm
(n = 22) and was greater than T4, which
was 13.6 N/mm (n = 22). These values
are similar to findings by Lee et al,32 with
initial mean stiffness greater at T5 (14.8
N/mm) than T4 (13.6 N/mm). Authors
have hypothesized that vertebral levels
higher in the thoracic spine may dem-
onstrate more PA movement and less PA
Downloaded from www.jospt.org at on December 7, 2015. For personal use only. No other uses without permission.
stiffness because they are located closer
to the highly mobile cervical spine.32,50
This may be due to vertebral geometry
and orientation between consecutive in-
tervertebral levels and localized muscle
action.32 This hypothesis is supported by
Copyright 2010 Journal of Orthopaedic & Sports Physical Therapy. All rights reserved.
this study, as the lowest premanipula-
tion stiffness values were at T3 (11.7 N/
mm, n = 19) followed by T4 and then T5.
Edmonston et al14 also identified lower
thoracic PA spinal stiffness values in
higher thoracic vertebral levels, with sig-
nificantly less stiffness at T4 compared to
T7. This has implications for therapists
when selecting thoracic vertebral levels
Journal of Orthopaedic & Sports Physical Therapy
to manipulate in symptomatic patients.
Therapists should consider that the stiff-
ness they palpate at higher levels in the
thoracic spine is usually less than at lower
levels in normal asymptomatic spines.
Therapist Selection of Spinal Levels
In clinical practice, therapists typically
identify patients or spinal levels with in-
creased spinal stiffness as more likely to
benefit from manipulation,1 and provide
treatment to reduce stiffness to poten-
tially improve thoracic spine symptoms.14
The current study provides some support
for targeting a specific spinal level for
manipulation. The association between
the magnitude of spinal stiffness and its
change following manipulation only oc-
curred at the level manipulated and not
at other levels. This suggests that any me-
chanical effects that might occur during
manipulation potentially affect the spi-
nal level where manipulation is applied
rather than multiple levels throughout
journal of orthopaedic & sports physical therapy|volume 40|number 11|november 2010|691
40-11 Campbell.indd 691
the spine.
The ability of physical therapists to
select a patients stiffest spinal level is
debatable. Landel et al28 reported that
therapists were unreliable in selecting the
stiffest lumbar segment, when stiffness
was defined by the movement observed
with dynamic magnetic resonance im-
aging. The authors suggested therapists
likely use their perception of resistance,
as well as movement, in selecting stiff
vertebral levels, as intertherapist reliabil-
ity for selecting the stiffest level was good.
Intertherapist reliability in selecting stiff
vertebral levels was also reported as good
when therapists assigned stiffness values
using a palpated reference standard.8
In the current study, the therapist
selected the stiffest of the 5 measured
vertebral levels in only 6 subjects (25%),
suggesting difficulty in accurately per-
ceiving differences in palpable stiffness
between consecutive vertebral levels in
individual subjects. Maher and Adams34
report that physical therapists can cor-
rectly distinguish an 11% difference in
stiffness. The differences in mean stiffness
between vertebral levels in this study were
approximately this magnitude (eg, mean
T5 premanipulation stiffness was 11%
higher than T4); however, the differences
between vertebral levels for an individual
might have been smaller. These small dif-
ferences might not be substantial enough
for a therapist to identify a spinal level as
palpably stiffer than the adjacent one.
Limitations
This study is limited to the immediate
short-term effect of PA spinal stiffness fol-
lowing thoracic manipulation in asymp-
tomatic people and cannot be generalized
to symptomatic patients. Symptomatic
patients are thought to have increased
PA spinal stiffness compared to normal
individuals, attributable to the presence
of pain and the contribution of muscle
spasm.32 There is evidence for the use of
spinal manipulation to reduce pain and
reflex muscle spasm,25 which may poten-
tially contribute to feelings of pain relief
following manipulation without any as-
sociated change in spinal stiffness.
The small sample size in this study
could be one reason that there was not
a statistical difference between prema-
nipulation and postmanipulation stiff-
ness. There are many factors that affect
spinal stiffness, and the sample size in the
current study might not have been large
enough to detect stiffness differences ac-
counting for all other possible factors.
Furthermore, individual differences in
stiffness premanipulation and postma-
nipulation for some subjects were small.
The age of the subjects in this study might
be one reason for this. Young people are
less likely to have degenerative changes,
such as osteoarthritis, which is related
to increasing age.48 Further, increased
age has been previously shown to be as-
sociated with greater cervical PA spinal
stiffness at C2, when measured with the
device used in the current study.45 As
subjects in the current study were young,
their thoracic spines might have been
relatively mobile, resulting in smaller
changes following manipulation.
Though there was a greater number of
subjects whose spinal stiffness decreased
after manipulation, there were some sub-
jects with higher measured stiffness fol-
lowing manipulation compared to their
premanipulation values. The reasons for
this are unknown, though some authors
propose that manipulative procedures
may potentially stimulate excitatory
behavior of the paraspinal muscles,16
possibly increasing resistance to PA
movement.32 Possible measurement er-
ror might be another factor contribut-
ing to the nonsignificant results. The
standard error of measurement for the
stiffness device is 0.83 N/mm, which
suggests that smaller differences in stiff-
ness were undetected. Lastly, neurophysi-
ological effects, such as the stimulation of
mechanoreceptors reducing muscle tone,
cannot be ruled out as contributors to po-
tential mechanical changes observed in
PA spinal stiffness.
Future Research
Future studies involving older subjects
10/20/10 1:00 PM
 [ research report
and symptomatic patients should be
conducted to determine if those subjects
have greater thoracic spine stiffness and
may subsequently demonstrate a great-
er reduction in stiffness following ma-
nipulation. In addition, future research
should investigate if there are potential
changes in paraspinal muscle tone or
neurophysiological mechanisms follow-
ing manipulation. These factors may con-
tribute to the reduction in pain that some
patients experience after manipulation, if
mechanical changes such as spinal stiff-
Downloaded from www.jospt.org at on December 7, 2015. For personal use only. No other uses without permission.
ness are minimal.
CONCLUSION
In asymptomatic individuals, spinal
stiffness does not appear to change fol-
Copyright 2010 Journal of Orthopaedic & Sports Physical Therapy. All rights reserved.
lowing manipulation using a PA thrust
in supine lying. However, stiffness differs
between spinal levels, and any potential
mechanical effects of manipulation ap-
pear to occur at the targeted level rather
than other levels. Greater postmanipu-
lation stiffness is associated with higher
premanipulation stiffness and with male
gender. These results suggest that, while
Journal of Orthopaedic & Sports Physical Therapy
the perceived benefits of thrust manipu-
lation may not be a consequence of spinal
stiffness changes, targeting treatment to
a specific spinal level may be beneficial.
KEY POINTS
FINDINGS: This study found that thoracic
spine stiffness was not statistically sig-
nificantly different following manipu-
lation in asymptomatic individuals.
However, observed changes in spinal
stiffness were associated with stiffness
values at the manipulated level only,
suggesting that any potential mechani-
cal effects are targeted to the level
manipulated.
IMPLICATION: The clinical practice of
targeting manipulation to specific spinal
levels appears worthwhile, though the
benefits of manipulation cannot be ex-
plained by stiffness changes in the spine.
CAUTION: Findings are limited to short-
term PA spinal stiffness changes in
young asymptomatic subjects and should
692|november 2010|volume 40|number 11|journal of orthopaedic & sports physical therapy
40-11 Campbell.indd 692
]
not be generalized to reflect changes in
patients with spinal symptoms.
ACKNOWLEDGEMENTS: The authors would like
to acknowledge Lucy Thomas, DipPhys,
GradDipAppSc(ManipPhty), MMedSc
(Phty) for assistance with data collection,
and the financial assistance provided by
The University of Newcastle Summer
Vacation Research Scholarship and the
Vice Chancellors Honors Scholarship
that supported Brad Campbell.
REFERENCES
1. Abbott JH, Flynn TW, Fritz JM, Hing WA, Reid
D, Whitman JM. Manual physical assessment
of spinal segmental motion: intent and valid-
ity. Man Ther. 2009;14:36-44. http://dx.doi.
org/10.1016/j.math.2007.09.011
2. Bialosky JE, George SZ, Bishop MD. How spinal
manipulative therapy works: why ask why? J
Orthop Sports Phys Ther. 2008;38:293-295.
http://dx.doi.org/10.2519/jospt.2008.0118
3. Brodeur R. The audible release associated with
joint manipulation. J Manipulative Physiol Ther.
1995;18:155-164.
4. Bronfort G, Haas M, Evans RL, Bouter LM. Ef-
ficacy of spinal manipulation and mobilization
for low back pain and neck pain: a systematic
review and best evidence synthesis. Spine J.
2004;4:335-356. http://dx.doi.org/10.1016/j.
spinee.2003.06.002
5. Caling B, Lee M. Effect of direction of applied
mobilization force on the posteroanterior
response in the lumbar spine. J Manipulative
Physiol Ther. 2001;24:71-78. http://dx.doi.
org/10.1067/mmt.2001.112568
6. Chansirinukor W, Lee M, Latimer J. Contribution
of ribcage movement to thoracolumbar postero-
anterior stiffness. J Manipulative Physiol Ther.
2003;26:176-183. http://dx.doi.org/10.1016/
S0161-4754(02)54131-2
7. Childs JD, Fritz JM, Flynn TW, et al. A clinical
prediction rule to identify patients with low
back pain most likely to benefit from spinal ma-
nipulation: a validation study. Ann Intern Med.
2004;141:920-928.
8. Chiradejnant A, Maher CG, Latimer J. Objective
manual assessment of lumbar posteroanterior
stiffness is now possible. J Manipulative Physiol
Ther. 2003;26:34-39. http://dx.doi.org/10.1067/
mmt.2003.3
9. Cleland JA, Childs JD, McRae M, Palmer JA,
Stowell T. Immediate effects of thoracic manipu-
lation in patients with neck pain: a randomized
clinical trial. Man Ther. 2005;10:127-135. http://
dx.doi.org/10.1016/j.math.2004.08.005
10. Cleland JA, Fritz JM, Whitman JM, Childs JD,
Palmer JA. The use of a lumbar spine manipula-
tion technique by physical therapists in patients
who satisfy a clinical prediction rule: a case se-
ries. J Orthop Sports Phys Ther. 2006;36:209-
214. http://dx.doi.org/10.2519/jospt.2006.2163
11. Conway PJW, Herzog W, Zhang Y, Hasler EM,
Ladly K. Forces required to cause cavitation
during spinal manipulation of the thoracic spine.
Clin Biomech. 1993;8:210-214. http://dx.doi.
org/10.1016/0268-0033(93)90016-B
12. Cramer GD, Tuck NR, Jr, Knudsen JT, et al.
Effects of side-posture positioning and side-
posture adjusting on the lumbar zygapophysial
joints as evaluated by magnetic resonance
imaging: a before and after study with ran-
domization. J Manipulative Physiol Ther.
2000;23:380-394. http://dx.doi.org/10.1067/
mmt.2000.108145
13. Descarreaux M, Dugas C, Lalanne K, Vincelette
M, Normand MC. Learning spinal manipula-
tion: the importance of augmented feedback
relating to various kinetic parameters. Spine
J. 2006;6:138-145. http://dx.doi.org/10.1016/j.
spinee.2005.07.001
14. Edmondston SJ, Allison GT, Althorpe BM, Mc-
Connell DR, Samuel KK. Comparison of ribcage
and posteroanterior thoracic spine stiffness:
an investigation of the normal response. Man
Ther. 1999;4:157-162. http://dx.doi.org/10.1054/
math.1999.0202
15. Edmondston SJ, Singer KP. Thoracic spine:
anatomical and biomechanical considerations
for manual therapy. Man Ther. 1997;2:132-143.
http://dx.doi.org/10.1054/math.1997.0293
16. Evans DW. Mechanisms and effects of spinal
high-velocity, low-amplitude thrust manipula-
tion: previous theories. J Manipulative Physiol
Ther. 2002;25:251-262.
17. Fernandez-de-las-Penas C, Palomeque-del-
Cerro L, Rodriguez-Blanco C, Gomez-Conesa A,
Miangolarra-Page JC. Changes in neck pain and
active range of motion after a single thoracic
spine manipulation in subjects presenting with
mechanical neck pain: a case series. J Manipu-
lative Physiol Ther. 2007;30:312-320. http://
dx.doi.org/10.1016/j.jmpt.2007.03.007
18. Forand D, Drover J, Suleman Z, Symons B, Her-
zog W. The forces applied by female and male
chiropractors during thoracic spinal manipula-
tion. J Manipulative Physiol Ther. 2004;27:49-
56. http://dx.doi.org/10.1016/j.jmpt.2003.11.006
19. Fritz JM, Whitman JM, Childs JD. Lumbar spine
segmental mobility assessment: an examination
of validity for determining intervention strate-
gies in patients with low back pain. Arch Phys
Med Rehabil. 2005;86:1745-1752. http://dx.doi.
org/10.1016/j.apmr.2005.03.028
20. Gangnet N, Dumas R, Pomero V, Mitulescu
A, Skalli W, Vital JM. Three-dimensional
spinal and pelvic alignment in an asymp-
tomatic population. Spine (Phila Pa 1976).
2006;31:E507-512. http://dx.doi.org/10.1097/01.
brs.0000224533.19359.89
21. Gross J, Fetto J, Rosen E. Musculoskeletal Exam-
ination. 2nd ed. Malden, MA: Blackwell Science,
Inc; 2002.
10/20/10 1:00 PM
 22. Herzog W, Kats M, Symons B. The effective
forces transmitted by high-speed, low-amplitude
thoracic manipulation. Spine (Phila Pa 1976).
2001;26:2105-2110; discussion 2110-2101.
23. Herzog W, Zhang YT, Conway PJ, Kawchuk GN.
Cavitation sounds during spinal manipula-
tive treatments. J Manipulative Physiol Ther.
1993;16:523-526.
24. Hoppenfeld S. Physical Examination of the
Spine and Extremities. New York, NY: Appleton-
Century-Crofts; 1976.
25. Katavich L. Differential effects of spinal ma-
nipulative therapy on acute and chronic muscle
spasm: a proposal for mechanisms and efficacy.
Man Ther. 1998;3:132-139.
26. Kenna C, Murtagh J. Back Pain and Spinal
Manipulation: A Practical Guide. Oxford, UK:
Downloaded from www.jospt.org at on December 7, 2015. For personal use only. No other uses without permission.
Biomech. 1993;8:302-306. http://dx.doi.
org/10.1016/0268-0033(93)90004-2
33. Lee M, Steven GP, Crosbie J, Higgs RJ. Variations
in posteroanterior stiffness in the thoracolumbar
spine: preliminary observations and proposed
mechanisms. Phys Ther. 1998;78:1277-1287.
34. Maher C, Adams R. A psychophysical evaluation
of manual stiffness discrimination. Aust J Phys-
iother. 1995;41:161-167.
35. Maitland GD, Hengeveld E, Banks K, English K.
Maitlands Vertebral Manipulation. 7th ed. Phila-
delphia, PA: Elsevier Butterworth-Heinemann;
2005.
36. Martinez-Segura R, Fernandez-de-las-Penas
C, Ruiz-Saez M, Lopez-Jimenez C, Rodriguez-
Blanco C. Immediate effects on neck pain and
active range of motion after a single cervical
loading. Man Ther. 2002;7:19-25. http://dx.doi.
org/10.1054/math.2001.0432
43. Shirley D, Hodges PW, Eriksson AE, Gande-
via SC. Spinal stiffness changes through-
out the respiratory cycle. J Appl Physiol.
2003;95:1467-1475. http://dx.doi.org/10.1152/
japplphysiol.00939.2002
44. Snodgrass SJ, Rivett DA, Robertson VJ. Calibra-
tion of an instrumented treatment table for
measuring manual therapy forces applied to
the cervical spine. Man Ther. 2008;13:171-179.
http://dx.doi.org/10.1016/j.math.2007.04.002
45. Snodgrass SJ, Rivett DA, Robertson VJ. Measur-
ing the posteroanterior stiffness of the cervical
spine. Man Ther. 2008;13:520-528. http://
dx.doi.org/10.1016/j.math.2007.07.007
46. Stoddard A. Manual of Osteopathic Practice.

Butterworth-Heinemann; 1997.
27. Kirstukas SJ, Backman JA. Physician-applied
contact pressure and table force response dur-
ing unilateral thoracic manipulation. J Manipula-
tive Physiol Ther. 1999;22:269-279.
28. Landel R, Kulig K, Fredericson M, Li B, Powers
CM. Intertester reliability and validity of motion
Copyright 2010 Journal of Orthopaedic & Sports Physical Therapy. All rights reserved.
high-velocity low-amplitude manipulation in
subjects presenting with mechanical neck pain:
a randomized controlled trial. J Manipulative
Physiol Ther. 2006;29:511-517. http://dx.doi.
org/10.1016/j.jmpt.2006.06.022
37. Neter J, Kutner MH, Wasserman W, Nachtsheim
CJ. Applied Linear Regression Models. 3rd ed.
2nd ed. London, UK: Harpercollins Publishers
Ltd; 1983.
47. Tuttle N, Barrett R, Laakso L. Posteroanterior
movements in tender and less tender locations
of the cervical spine. Man Ther. 2009;14:28-35.
http://dx.doi.org/10.1016/j.math.2007.09.003
48. van Saase JL, van Romunde LK, Cats A, Vanden-

assessments during lumbar spine accessory
motion testing. Phys Ther. 2008;88:43-49.
http://dx.doi.org/10.2522/ptj.20060179
29. Latimer J, Goodsel MM, Lee M, Maher CG,
Wilkinson BN, Moran CC. Evaluation of a new
device for measuring responses to postero-
anterior forces in a patient population, part 1:
reliability testing. Phys Ther. 1996;76:158-165.
30. Latimer J, Lee M, Adams R, Moran CM. An
investigation of the relationship between low
back pain and lumbar posteroanterior stiffness.
Journal of Orthopaedic & Sports Physical Therapy
Chicago, IL: McGraw-Hill/Irwin; 1996.
38. Palmer ML, Epler ME. Fundamentals of Mus-
culoskeletal Assessment Techniques. 2nd ed.
Philadelphia, PA: Lippincott-Raven Publishers;
1998.
39. Schiller L. Effectiveness of spinal manipulative
therapy in the treatment of mechanical thoracic
spine pain: a pilot randomized clinical trial. J
Manipulative Physiol Ther. 2001;24:394-401.
http://dx.doi.org/10.1067/mmt.2001.116420
40. Schiotz EH. Manipulation Past and Present.
broucke JP, Valkenburg HA. Epidemiology of os-
teoarthritis: Zoetermeer survey. Comparison of
radiological osteoarthritis in a Dutch population
with that in 10 other populations. Ann Rheum
Dis. 1989;48:271-280.
49. Viner A, Lee M, Adams R. Posteroanterior
stiffness in the lumbosacral spine. The correla-
tion between adjacent vertebral levels. Spine
(Phila Pa 1976). 1997;22:2724-2729; discussion
2729-2730.
50. White AA, Panjabi MM. Clinical Biomechanics

J Manipulative Physiol Ther. 1996;19:587-591. London, UK: Heinemann Medical; 1975. of the Spine. 2nd ed. Philadelphia, PA: J.B. Lip-
31. Latimer J, Lee M, Adams RD. The effects of high 41. Shirley D. Manual therapy and tissue stiffness. pincott Company; 1990.
and low loading forces on measured values of In: Boyling JD, Jull G, eds. Grieves Modern Man-
lumbar stiffness. J Manipulative Physiol Ther. ual Therapy: The Vertebral Column. Edinburgh,

@ more information
1998;21:157-163. UK: Churchill Livingstone; 2004:381-390.
32. Lee M, Latimer J, Maher C. Manipulation: 42. Shirley D, Ellis E, Lee M. The response of
investigation of a proposed mechanism. Clin posteroanterior lumbar stiffness to repeated www.jospt.org

PUBLISH Your Manuscript in a Journal With International Reach

JOSPT offers authors of accepted papers an international audience. The
Journal is currently distributed to the members of APTAs Orthopaedic
and Sports Physical Therapy Sections and 18 orthopaedics, manual therapy,
and sports groups in 17 countries who provide online access as a member
benefit. As a result, the Journal is now distributed monthly to more than
30 000 individuals around the world who specialize in musculoskeletal and
sports-related rehabilitation, health, and wellness. In addition, JOSPT
reaches students and faculty, physical therapists and physicians at more
than 1,400 institutions in 55 countries. Please review our Information for and
Instructions to Authors at www.jospt.org and submit your manuscript for
peer review at http://mc.manuscriptcentral.com/jospt.

journal of orthopaedic & sports physical therapy|volume 40|number 11|november 2010|693

40-11 Campbell.indd 693 10/20/10 1:00 PM