Upsala J Med Sci 100: 151-160,1995

A Clinical Method for Measuring the Distribution of
Segmental Flexion Mobility in the

Cervico-Thoracic Spine
Staffan Norlander, RPT,’ Ulnka Aste-Norlander, RPT: Bengt Nordgren, MD, PhD3

and Bo Sahlstedt, MD, PhD3
’Research Foundation for Occupational Safety and Health in the Swedish Construction Industry,2 US Fysioterapi, Trosa

and Departmentv of ‘Rehabilitation Medicine and Diagnostic Radiology, Akademiska Sjukhuset,
Uppyala Universir): Uppsala, Sweden

ABSTRACT

The aim of this study was to evaluate the validity and the repeatability of a new technique to assess
segmental flexion mobility in the cervico-thoracic spine between segments C7 and T5. The new
technique is referred to as the Cervico-Thoracic-Ratio (the CTR-technique). The radiological
evaluation of skin distraction measurements showed that validity was high for the CTR-technique.
A high correlation between vertebral flexion mobility and skin distraction was recognized individu-
ally and for the whole group. The evaluation of repeatability was found to be high for intratester and
fair for intertester repeatability. The CTR-technique may become a valuable complement to other
method? for assessing segmental flexion mobility in patients suffering from neck-shoulder pain in
clinical practice.

INTRODUCTION

A new technique has been described tomeasure the segmental flexion mobility in thecervico-thoracic
junction and the upper thoracic spine (7). The measuring technique is referred to as the Cervico-
Thoracic-Ratio (the CTR-technique). The CTR-technique describes what is defined as relative
flexion mobility (CTR%). Relative flexion mobility is a calculated ratio based on absolute values of
skin distraction between C7 - T5. Absoluteflexinn mobility is defined as the measured alteration in
cm between 3 cm interdistant skin markings marked from C7 down t o T 5 and measured with a tape
measure after a maximal forward flexion of the trunk and neck from an upright posture. The distance
of 3 cm marked, in the upright posture, has been used as the definition of one motion segment, as the
height of one disc and one thoracic vertebral body is approximately 3 cm, according to (5).

The validity of the CTR-technique is dependent on an individual and strong reiationship bezween
vertebralflexion mobility and skin distraction in the motion segments between C7 and T5, since the
method is meant to be used for individual assessment of flexion mobility. Several attempts have been
made to establish the relationship between spinal mobility and different methods for cliiiical
examination. According to (2) and (9) clinical examinations and radiographic measurements showed
high validity for measurement of lumbar curvature during stance and trunk forward flexion.

151

According to (1) and (10) different clinical examinations showed poor validity compared with
measurements from radiographic pictures. I n conclusion, validity differs for different instruments
and methods. The CTR-technique has to be evaluated i n order to become areliable method for clinical
examination of the segmental flexion mobility.

Repeatubiliry is defined as the capability to repeat a measurement. Many clinical methods for
examination of spinal mobility are known to show low repeatability. Different factors affect the
repeatability, for example if the mobility is tested passively or actively, or how long the time interval
is between repeated measurernents. According to (3)the best repeatability is obtained if the mobility
is tested actively and with as short time interval as possible between measurements, the classic " ~ e s
retest design". In the CTR-technique mobility is tested actively. Repeatability is also known to be
higher when it is perfomied by one tester, intratester repeutubility, compared with measurements
done by two testers, intertester repeutuhility. In order to get acceptance of the CTR- technique the
repeatability also has to be evaluated. The aim of this study is to evaluate the validity a n d the
repeatability of the CTR- technique as a method to be used in clinical practice for assessment of the
segmental flexion mobility between C7 and TS.

MATERIALS AND METHODS

Radiological evaluation of validity

The validity of the CTR-technique is dependent on a n individual and strong relationship between
vertebral flexion mobility and skin distraction i n motion segments between C7 and TS. In order 10
study the individual relationship six different vertebral angles C7 to TS were evaluated against six
corresponding skinmarkings for each subject. The analysis of relationship was also evaluated for the
whole group of 42 different vertebral angles versus 42 corresponding skinmarkings. Seven male
subjects with pain in the neck-shoulder region, mean age 40.3, (SD16.0) years participated i n the
evaluation of validity. The evaluation was only done on male subjects, since previous studies did not
show any significant differences between female and male subjects with reference to the degree of
skin distraction, during forward flexion measured according to the CTR-technique(7).

Procedure. Six small metal markings were glued onto the skin with 3 cm intervals, according to the
marking procedure described i n the CTR-technique. The markers were glued with the subject in an
upright sitting posture and the upper marking p u t over the most prominent part of the spinous process
of C7. Lateral radiographs were used to obtain overlay measurements of the alteration of vertebral
angles and of skin marker (Fig. 1). The first radiograph was taken with the subject i n an upright sitting
posture and the second in a maximal flexed sitting posture. Since the spinous processes could not be
demonstrated in the thoracic region without tomography, the angle between the vertebral endplates
were used as the independent variable. An aluminium wedge was usedin order to equalize the contrast
differences and to visualize the metal markers on the skin. The angles between the endplates and the
distances between the markers were measured on the films with the patient i n upright position a n d

152

i n maximal forward flexion. The T6-vertebra was used as a reference vertebra, the cumulative angles
of C7 down to TS were measured with a gauge. The diagonal alteration of metal skin markers M 1 -
M6 were measured with a pak of compasses and a ruler (Fig. 1). The measurements were done only
once.

Figure 1 Lateral radiographs were used Lo obtain overlay measurements of Lhecumulative vertebral angles (c7 - TS)
and the diagonal alterations of the metal markings (M1 - M6).

Statistical analysis. The determination of the relationship between skin distraction and vertebral
flexion mobility was done by deciding the highest regression coefficient for five different regression
models. Both the relationship for the individual and the whole group was evaluated for each model.
A computerprogrmi (Quest) calculated the equation for linear (Y=A+B*X), exponential ( Y =Aiexp
(B*X)), power (Y=A*XAB), logarithmic (Y=A+B*ln ( X ) ) and polynomial models ( Y = A - B , * X +
B,* X2). Vertebral angles were used as the independent variable, and skinmarkers as the dependent
variable. Breakdowns with one-way anova were also used to describe the mean values of the
dependent variable skinmarkers as a function of the independent variable vertebral angles.

Evaluation of repeatability

The evaluation of repeatability was done for intra- and intertester repeatability. Tests were done by
two investigators, two trials each. The evaluation was done with a test-retest design on 26 male
subjects, mean age 41.2, (SD 9.3) years.

Procedure . On arrival the subject was instructed to sit in a chair. The first investigator put the
markings on the subject according to the examination procedure described in the CTR-technique, and
the subject was asked to flex forward as much as possible. The investigator measured and noted all
the five alterations. After that the markings were erased. The same procedure was repeated by the
second investigator. Finally a second trial was repeated by both investigators, altogether four trials.

12-950246 153

Statistical analysis. In the evaluation of repeatability a sign-test was used to reveal byslernatic errors.
Random errors were evaluated by calculating t h e measuring precision and the relative measuring
error. Both absolute and relative flexion mobility was evaluated. The precision was calculated as the
pooled standarddeviations of the differences between trialsor investigators, divided by the extracted
square root of two. The relative measuring error, the coefficient of variation (CV) was calculated as
the standard deviation divided by the mean, times 100. Breakdown with one way anova was used to
analyse the degree of conformity between intra- and intertester repeatability expressed as R square
and p-values.

RESULTS

Evaluation of validity

The results of the analysis of the relationship between vertebral flexion mobiIity and skin distraction
show that the validity of the CTR-technique was individually very high. (Table 1). The degree of
relationship vaned between different regression models, which implies an individual variation of
spinal flexion mobility. The polynomial model showed the highest degree of individual relationship
in five subjects, the logarithmic model in two subjects and the power model and the linear model
showed equally high values as the polynomial model in one subject each. The different r2 values
ranged between 0.68-0.98, which is a very high degree of explanation and all seven subjects showed
a statistically significant relationship in at least one of the models (Table 1).

Table 1 Results of individual relationships between dependent variable skin distraction and independent variable
vertebral angles.

I Regression models

0.68
0.1 1
0.91

p-value

0.08
0.002
0.004
0.003
0.04
0.52
0.004

Exponential

0.89 0.006
0.83 0.13

0.56
0.07
0.97 0.001

Power i;i--I...I- Logarithmic +- Polynomial - rz p-value .~
0.6X 0.04
0.95 0.002
0.88 0.007
0.93 0.003
0.44 0.14
0.00 1.0
0.81 0.04

0 61 0.07
0.95 0.002
0 9 3 0003
0.96 0.002
0.56 0.09
0.01 0.81
0.70 0.08

0.68 0.18
0.96 0.008
0.92 0.023
0.94 0.012
0.82 0.07
0.91 0.03
0.98 0.003

For the whole group the polynomial regression model showed the strongest relationship between
vertebral flexion mobility arid skin distraction, r2= 0.44 arid p<0.001. (n=42) (Fig. 2). The linear
model showed r2= 0.39, (p<O.001), (n=42), the exponential model ?= 0.27 (p<O.OOI), (n=42), the
power model r2= 0.19 (p<0.005), (n=42) and the logarithmic model showed r2 = 0.28 (p<0.001),
(n=42). Themean cumulativevertebral angleforC7 was 18" and the alteration of the AM1 skinmarker
was 5.3 cm. The cumulative flexion mobility for TI and AM2 was 13" versus 3.8 cm. T2 and AM3
was 8.1' versus 3.0 cm. T3 and AM4 was 5.9" versus 2.5 cm. T4 and AM5 was 5 . I ' versus 2.1 cm.
T5 and AM6 was 4.6" versus 1.8 cm (Fig. 2).

154

7 -, I

y = 2 2525 - 0 UO81 x + 0 0078 xi

6 I

Figure 2 Regression model describing thc highest relationship between vertebraiangles (C7 - T5) and alteration of
skin markings (AM1 - AM6) (r = 0.66, p<O.Wl) (n = 32). Mean values (& SD) for studied variable>.

The breakdowns with one- way anova analysis showed a relationship similar to the polynomial model
(Table 2) between vertebral angles and skin distraction, r2 was 0.41 and p-value 0.001. Increasing
vertebral angles showed increasing skin distraction.

Table 2 Thc breakdown with one-way anova analysis showcd a similar rclationship between vertebral angles and
alterations of sktnmarkings, rZ = 0.413 and p< 0.001 (n = 42).

Vertebral ... Skinmarkings (cm)
angles ( )" X SD Min M a x
0 - 6 2.335 0.916 0.600 - 3.200
6 - 12 2.883 1.389 1.300 - 5.800

12 - 18 3.622 1.645 1.300 - 5.800
I 8 - 24 5.300 0.781 4.400 - 5.800
24 - 30 6.700 - 6.700 - 6.700

4 Number crf,ioints 17
12
9
3
1

Total 3.083 1.546 0.600 - 6.700 42

Evaluation of repeatability

Absoliiteflex.xionmobilily. Theresults of theevaluation showed that repeatability was high fc~rabsolute
flexion mobility, for both intra- and intertester repeatability, which was important, as the values of
relative flexion mobility exclusively depended on the values of absolute flexion mobility. The
absolute flexion mobility was measured in cm according to the description of the CTR-technique
from motion segment C7 down to T5. The inrrutester repeatability showed no systematic error
between the first and second trial either for tester 1 or tester 2, when absohte flexion mobility was

155

evaluated for all motion segments. The sign test did not show any significant differences between
trials for the two testers (Table 3, I* and 2*).The intertester repeatability, comparing tester 1 with
tester 2 showed a small, but significant systematic measuring error between testers in both trials and
in all five motion segments, (Table 3,3* and 4 9 . Tester 2 systematically measured a 2-3 mm shorter
absolute flexion mobility for all motion segments, The random errors expressed as measuring
precision (mm) and as relative measuring error CV showed that repeatability was high for both intra-
and intertester repeatability (Table 3, 1* and 4*). In general the precision is higher the shorter the
distance measured , while the CV is greater. The CV for intratester repeatability ranged from 2.1-
4.8% for tester 1 and from 1.9 - 4.4% for tester 2 for the different motion segments (Table 3, I* and
2*). The CV for inlertesterrepeatability ranged from 2.4- 5.7% for trial 1 and from 1.8-4.1 % for trial
2 for the different motion segments (Table 3,3* and4*). The breakdown with one way aiiova analysis
describing the relationship between repeated trials for absolute flexion mobility showed a very high
degree of conformity, r2values ranged 0.995 - 0.998 and p-values was p<O.OOI, for both intra- and
intertester repeatability.

Table 3 Results Irom evaluation ofrcpedtability of absolute flexion [nobility (n = 26), 1* and 2" = inwitester, 3*
and 4* = intertester repeatability.

Absolute flexion
mobility (cm)
X SD

3.8 0.2
3.6 0.2
3.8 0.2
3.6 0.2

7.3 0.3
7.1 0.4
7.3 0.3
7.1 0.3

10.6 0.4
10.5 0.5
10.6 0.4
10.4 0.5

14.0 0.5
13.8 0.6
14.0 0.5
13.7 0.5

17.4 0.5
17.1 0.6
17.3 0.5
17.1 0.6

Motion
segment Trial Tester

Random errors Systematic
Measuring Measuring errors
precision (mm) error ( C V % ) I Sign test

I * 1.8 4.8
2* 1.6
3* 2.1 5.7
4* 1.5 4.1

I* 2.7 3.6
2* 2.4 3.4 11s
3* 3.1 4.3 ns
4* 2.5 3.5 4" 0.01

1* 2.7 2.5 ns
' 2* 2.5 2.4 2* 11s

3' 4.2 4.0 :1* ns
4* 2.8 2.6 J* 0.0 I

I * 3.1 2.3 ' 1 * n.:
2" 3.0 2.2 2* ns
3* 3.6 2.6 3" 0.05

4* ns 4* 3.1 2.2
2.1 I * ns I * 3.6

2* 3.3 1.9 2* n s
3* 4.1 2.4 3* 0.02

4* ns 4" 3.0 1.8

~ ~ _ _ _

~~~

C 7 - T 1 1 1
2 2
3 1
4 2

c 7 - T 2 1 1
2 2
3 1
4 2

C 7 - T 3 1 1
2 2
3 1
4 2

c 7 - 7 - 4 1
2
3
4

C 7 - T 5 1
2
3
4

Total
1 * lntratestcr

2* Intratester

3* Intertester

4* Intertester

repeatability tester 1 10.60 4.8 2.8 3. I n b

repeatability tester 2 10.40 4.8 2.6 2.9 11s

repeatability trial 1 10.51 4.8 3.4 3.8 0.01

repeatability trial 2 10.49 4.8 2.6 2.8 0.01
_ _ _ _

Relative,flexion mobiliry. Relative flexion mobility expressed as CTR% is based on two values of
absolute flexion mobility. This makes relative values afflicted with a somewhat greater measuring
error. The intratester repeatability showed no systematic errors between the first and second trial
either for tester 1 or tester 2 when relative flexion mobility was evaluated for all motion segments.
The sign test did not show any significant differences between trials for the two testers (Table 4, 1 *
and 2*). Theintertesrclrrepeatability, comparing tester 1 with tester 2 , showed a significant systematic
measuring error between testers i n both trials for motion segment C7-T 1 and T4 - TS (Table 4 , 3 * and
4*). The systematic measuring error in absolute flexion mobility between testers resulted i n a
calculated CTR value which in average is 0.7% lower for motion segment C7-Tl and 0.4% higher
for motion segment T4 - T5 for tester 2 compared with tester 1. The random errors expressed as
measuring precision (%) and as relative measuring error CV showed that repeatability was high for
intratester repeatability and fair for intertester repeatability (Table 4, 1" and 4*). In general the
precision is higher in motion segments C7-T1 andT4 - 'I'S arid so is also the CV compared with motion
segments i n between. The CV for intratester repeatability ranged from 2.9-3.9%; for tester 1 and from
2.9 - 4.4% for tester 2 for the different motion segments (Table 4, 1 * and 2*) The CV for intertester
repeatability ranged from 5.4- 7.7% for trial 1 and from 3.2-7.7% for trial 2 for the different motion
segments (Table 4, 3" and 4").

Table 4 Resultsfromevaluationofrc~catabilitvofrelativeflexionmobilit\;(n=26), I * m d 2 * =intriltcster,3* and
4* = menester repeatability

I Relative flexion 1 Random error3
Motion mobilitj ( C I R % ) Measuring Measuring

segment Trial Tester X S D

C 7 - 1 ' 1 1 1
2 2
3 1
4 2

1'1 -1'2 1 1
~~~

2 2
3 1
4 2

T 2 - T 3 1 1
2 2
3 1

2 2
3 1

2 2
3 I
4 2

2 I .x 1.2
21.2 1 .0
22.1 O.Y
21.3 1.1
20.0 0.9
20.4 1.3
20.0 0.8
20.0 1.4
19.3 1 .o
19.8 I .3
19.6 1 .o
19.6 1.7
19.7 1.6
19.4 1.5
19.5 1 .s
19.3 1.3
19.3 1.7
19.3 1 .0
18.8 1.4
19.6 1 .0

Total
1 * Intratester

2" Intratester

'3* Intertester

4* Intertester

repeatability tester 1

repeatability tester 2

repeatability trial 1

repeatability trial 2

20.0 1.5

20.0 I .3

20.0 1.2

20.0 1.3

precision ('70) error ( C V 7 o )
I * 0.9 39
2' 0.8 3.5
3* 1.3 6.3
4* 0.7 3.2
I* 0.6 3.1
2* 0.9 4.4
3* 1.1 5.4
4* 1.3 6.4
1 * 0.6 2.9
2* 0.8 3.8
3* 1 . 1 6.0
4) 1.5 7.7
I * 0.7 3.1
2* 0.7 3.1
3* 1.5 1.7
4* 1.3 6.8
I * 0.7 3.7
2* 0 6 2.9
3* 1 4 7 0
4* I 0 5 1

0.7 3.5

0 . X 3.7

1.3 6.5

1.2 5.9

Systematic
errors
Sign test

l x n b
2" ns
3 , 0.02
4* 0.01
I* ns
2* 11s
3* ns
4* ns
I * 11s
2* ns
1* ns
4" ns
1 * ns
2* 11s
?* ns
4* ns
I* ns
2* ns
3* ns
4* 0.02

- *

~~ -

~ ~~

~~~~ ~

0.01

157

The breakdown with one way anova analysis, describing the relationship between repeated trials for
relative flexion mobility, showed a very high degree of conformity. Intratester repeatability showed
a r 2=0.79 (p<O.001) fortester 1 andar2 =0.68 (p<0.001) fortester 2. Intertesterrepeatability showed
a r2= 0.38 (p<0.001) for trial 1 and a 1-2= 0.58 (p< 0.001) for trial 2. Not in any motion segment did
the variation of repeatability significantly exceed the limits for what is defined as ordinary mobility
i n the classification model, described by (7), (Fig. 3).

Mobility Profile

Motion segment
~~

Figure 3 The resuits or the iiiua- and inrerterrer repcatability evaluation shown in thc mobility prof1 lz. The variation
betwccn trials and testers is kepi w i t h i n the limils for ordinary mobility.

DISCUSSION

The results of this study have shown that the distribution of segmental flexion mobility i n the cervico-
dorsal region can be examined with the CTR-technique. The validity of skin distraction measure-
ments in the cervico-dorsal spine was found to be individually very high. In this study it was necessary
to use conventional X-ray technique to evaluate the relationship between skin distraction and
vertebral motion. It was also necessary to choose the vertebral endplates to study the alteration of
vertebral angles since the spinous processes could not be clearly identified in this difficult region.
Great efforts were made tooptimize the technique. The number of subjects was limited to seven, since
the relationship between individual vertebral flexion mobility and skin distraction was found to be
convincingly high for our application. The aim with the CTR-technique was assessinent of joint
flexion mobility, from a clinical point of view. The aim was not to decide the exact vertebral angle
i n degrees. Further more the method has shown an important clinical application as an instrument to
identify dysfunction in the cervico-thoracic junction correlating to neck-shoulder pain (8). The
evaluation of validity was done on patients suffering from neck-shoulder pain and the evaluation of
repeatability was done on healthy subjects. This was not a problem, since clinical trial has shown that

158

the CTR-method can distinguish between different flexion mobility in individuals with and without
increased risk for neck-shoulder pain (8).

The intratester repeatability demonstrated high repeatability and was not impaired by any systematic
measuring error.The intertester repeatability demonstrated fair repeatability for both absolute and
relative flexion mobility. The measuring procedure was impaired by a sinall but systematically
measuring error for both absolute and relative flexion mobility, values between testers showed
slightly different characteristics. It was most likely that the marking procedure of the 3 cm interdistant
penmarkings contributed to the systematic measuring error. However, for absolute flexion mobility,
the relative measuring error must be considered small, only in one trial and for one motion segment
did the CV exceed 5 % (Table 3). As tester 2 systematically measured a 2-3 mm shorter absolute
flexion mobility for all motion segments, the calculation of the CTR resulted in a value which on an
average was 0.7% lower for motion segment C7-Tl and 0.4% higher for motion segment T4 - TS for
tester 2 compared with tester 1. This increased CV to 7.7% as the highest for relative flexion mobility.
This variation in intertester repeatability for relative flexion mobility was, however. not greater than
i t was comprised within the limits for what was defined as ordinary mobility in the classification
model described by (7). Consequently the systematic measuring error for intertesterrepeatability was
less than what was defined as ordinary variation between individuals. As the CTR was a ratio between
two absolute flexion mobility values, it was quite obvious that the relative value was impaired by a
somewhat greater measuring error compared with absolute values. If the precision in absolute values
could be improved by approximately 1 mm i n each motion segment, t h e CV would decline and nor
exceed the 5% limit for relative calculations. This could most likely be achieved by using an
instrument with fixed 3 cm interdistances for marking instead of an ordinary tape measure.

Only a few studies have used the coefficient of variation (CV) of their data to describe repeatability
and to our knowledge nobody has studied measurements in the thoracic spine according to the C l R -
technique. The present study was therefore difficult to compare with other studies. However, some
studies must be commented on. Gill et a1 (4) found that the intratester variation was 1.5% for the low
backmeasurements using the modified Schober technique, with the subject examined i n a f u l l y flexed
sittingposture. Merittetal(6)founda6.6% intratestervariation with the modified Schobertechnique,
examined in a fully flexed standing posture. Thus, also other skin distraction methods have shown
values of coefficient of variation of the same magnitude.

REFERENCES

1. Burdett, R . , Brown, K.& Fall, M.: Reliability and validity of four instruments for measuring
lumbar spine and pelvic positions. Phys Ther 66: 677-684, 1986.

Fint, MM.: Lumbar posture: A study of roentgenographic measurements and the influence of
flexibility and strength. Research Quarterly 34: 15-20, 1963.

Gajdosik, R.& Bohannon, R.: Clinical measurements of range of motion. Phys Ther 67: 1867-
1872. 1987.

159

5 .

10.

Gill, K., Krag, MH., Johnson, CB., Haugh, LD.& Pope, MH.: Repeatability of four clinical
methods for assessment of lumbar spinal motion. Spine 13: 50-53, 1988.

Kapandji, IA.: The physiology of the joints. Volume three. The trunk and the vertebral
column. Churchill Livingstone pp 38-39, 1970.

Meritt, JL., McLean, TJ., Erickson, RP.& Offord, Kp.: Measurement of trunk tlexibility in
normal subjects. Reproducibility of three clinical methods. Mayo Clin 61: 192-197, 1986.

Norlander, S., Aste-Norlander, U . , Nordgren, B.& Sahlstedt, B.: A clinical method for
measuring segmental flexion mobility i n the cervico-thoracic spine and a model for classifica-
tion. The cervico-thoracic ratio, (in press), Scand J Rehab Med. 1995.

Norlander, S., Aste-Norlander, U., Nordgren, B.& Sahlstedt, B.: Mobility in the cervico-
thoracic junction an indicative factor i n neck-shoulder pain. A cross-sectional study, (submit-
ted), 1995.

Troup, JDG., Hood, CA.& Chapman,AE.: Measurernentsof the sagittal mobility of the l u m b a r
spine and hips. Ann Phys Med 9: 308-321, 1968.

Stokes, IAF., Bevins, TM.& Limn, RA.: Back surface cutwture and measur-enienr of lumbar-
spinal motion. Spine 12: 355-366, 1987.

Address for offprints:

Staffan Norlander
Bygghalsans forskningsstiftelse
Svardvagen 25 A
S-182 33 DANDERYD, Sweden

160